REVIEWS

GROWTH OF THE PLANT

CELL WALL
Daniel J. Cosgrove
Abstract | Plant cells encase themselves within a complex polysaccharide wall, which
constitutes the raw material that is used to manufacture textiles, paper, lumber, films,
thickeners and other products. The plant cell wall is also the primary source of cellulose, the
most abundant and useful biopolymer on the Earth. The cell wall not only strengthens the plant
body, but also has key roles in plant growth, cell differentiation, intercellular communication,
water movement and defence. Recent discoveries have uncovered how plant cells synthesize
wall polysaccharides, assemble them into a strong fibrous network and regulate wall
expansion during cell growth.

PROTOPLASM A tree’s leaves may be ever so good, some of the largest plant cells are XYLEM vessels (FIG. 1b),
The contents of living cells,
So may its bark, so may its wood which can increase in volume >30,000-fold from their
including cytoplasm and
Leaves Compared with Flowers, Robert Frost MERISTEMATIC initials. Similarly, hair cells on the surface of
nucleus.
young cotton seeds elongate 1000-fold before maturity.
CREEP Something there is that doesn’t love a wall Such growth is accomplished through enlargement of
Slow, time-dependent, the cell volume owing to water uptake into the VACUOLE
irreversible extension, in which
Mending Wall, Robert Frost
the microfibrils and associated
and irreversible extension of the pre-existing cell wall.
matrix polysaccharides slowly Without cell walls, plants would be pliant piles of Simultaneously, new polymers are integrated into the
slide within the wall, therefore PROTOPLASM, more like slime moulds than the stately wall to prevent it becoming thinner and weaker.
increasing its surface area. trees and other greenery that grace our planet. Plants In this article, I focus on recent progress in under-
comprise of ~35 cell types, each of which is distinc- standing how the plant cell wall grows, both in terms
XYLEM
A tissue that comprises a group tive in its size, shape, position and wall characteristics of the synthesis and integration of new polysaccharides
of specialized cells that are (FIG. 1a,b). In growing cells, the wall is typically a thin, into the wall and the expansion of the polysaccharide
involved in the transport of flexible layer (0.1–1 µm) that consists primarily of com- network. For many years, the enzymes that synthesize
water and solutes in vascular plex polysaccharides and a small amount of structural the wall polysaccharides remained elusive. However,
plants. Mature xylem vessels
essentially contain only the cell
proteins. The veil-like thinness of the wall can be visu- many of the genes responsible have been recently
wall. alized by light microscopy, whereas electron micro- identified, and we are now beginning to elucidate the
scopy reveals its fibrous character at the nanometer molecular machinery that assembles sugars into the
scale (FIG. 1c,d). Despite its thinness, the wall forms a complex polymers that comprise the cell wall. Similarly,
strong network that functions like a corset, compress- for many years the existence of ‘WALL LOOSENING’ enzymes
Department of Biology, ing and giving shape to the protoplast within. was proposed, without solid evidence concerning
208 Mueller Laboratory, Plant cells grow by expanding their cell walls through their nature. Now, strong contenders for this proc-
Penn State University, a process of controlled polymer CREEP. Because the cells ess have emerged and provide novel ideas about the
University Park, are tightly glued together through their cell walls, cell biochemical underpinnings of wall expansion. In this
Pennsylvania 16802, USA.
e-mail: migration is not possible and plant morphogenesis is review, I discuss only the PRIMARY PLANT CELL WALL and
[email protected] mostly a matter of localized cell division and selective not the SECONDARY CELL WALL, which is deposited after
doi:10.1038/nrm1746 cell enlargement. Such enlargement can be impressive: cells cease enlargement and often has a distinctive

850 | NOVEMBER 2005 | VOLUME 6 www.nature.com/reviews/molcellbio

 REVIEWS

Epidermal hair Phloem and have several functions3,4. They form hydrated gels
a Epidermis Cambium Xylem Cortex b that push microfibrils apart, easing their sideways
slippage during cell growth, while also locking them
in place when growth ceases. They are important
Xylem vessel
determinants of wall POROSITY and wall thickness and
they glue cells together in an adhesive layer called the
5
MIDDLE LAMELLA . Pectins are primary targets of attack by
invading microbes and their breakdown products func-
tion as potent elicitors of plant-defence responses.
Cellulose and MATRIX POLYSACCHARIDES are made by
distinctive pathways (FIG 2; see below). Cellulose is
synthesized by large membrane complexes6–8 which
extrude a microfibril from the cell surface, similar to
a spider’s thread. By contrast, matrix polysaccharides
are synthesized in the Golgi apparatus and packaged
into tiny vesicles that fuse with the plasma membrane
c d and thereby deliver their cargo to the wall. Matrix
polysaccharides then become integrated into the wall
network by physical interactions, enzymatic ligations
and crosslinking reactions. Unlike cellulose micro-
fibrils, newly secreted matrix polysaccharides can
diffuse some distance into the cell wall9, aided by cell
TURGOR PRESSURE, which stretches the cell wall, increases
its porosity and provides an energy gradient to drive
polymers into the wall10.

Cellulose synthesis. Plant cellulose synthase (CESA)

genes were identified in the late 1990s through
Figure 1 | Photographic depictions of plant cell walls. a | A cross-section of the stem of molecular and genetic studies11,12. In the model plant
Zinnia elegans illustrates the diverse cell sizes, shapes and cell-wall types. Note the tiny Arabidopsis thaliana, the CESA family contains ten
meristematic cells in the cambium, versus the huge cells that comprise the epidermal hairs and genes, which are expressed in different tissues and cell
the cortex. b | A scanning electron micrograph of a maize root cross-section129 illustrates the types. Growing evidence from genetic experiments
impressive ability of cells to enlarge. Note the huge xylem vessel. All of the cells in this cross- and gene-expression analyses indicates that three dif-
section originated from tiny meristematic initials. c | Micrograph of onion epidermal cell wall130.
ferent CESA genes are normally required to make a
This image, which was obtained by differential interference microscopy of wall fragments,
reveals the thinness and delicacy of the primary cell wall. d | A scanning electron micrograph of functional cellulose-synthesizing complex13–16 and that
the growing cell wall from cucumber hypocotyls50. Note the ordered fibrillar texture, which is the different sets of genes are involved in the formation of
result of the ordered arrangement of cellulose microfibrils, which are coated with a layer of the primary and secondary wall. For example, CESA1,
matrix polysaccharide. Part a of the figure is reproduced, with permission, from REF.128 © The CESA3 and CESA6 are required for biosynthesis of the
American Society of Plant Biologists (1995). Part b of the figure is reproduced, with permission, primary wall12,17, whereas CESA4, CESA7 and CESA8
from REF.129 © Department of Scientific and Industrial Research, New Zealand (1972). Part c of
are required to form secondary walls18.
the figure is reproduced, with permission, from REF.130 © The Company of Biologists (1990).
Part d of the figure is reproduced, with permission, from REF.50 © Blackwell Publishing (2005).
CESA proteins are embedded in the plasma mem-
brane in hexameric arrays called particle rosettes6
(FIG. 3a). The smallest subunits that are visible by electron
microscopy are thought to consist of six CESA proteins
MERISTEM
composition and organization. Not all cells have that are encoded by three genes (FIG. 3b,c). The assembly
Region of rapid cell division on
a plant; it is where cell initials secondary cell walls, which are made by cells that of CESA subunits into hexamers is not well understood,
(or ‘stem’ cells) are maintained require great mechanical strength and structural but is thought to require CESA dimerization, which is
and organogenesis starts. Root reinforcement. mediated by two zinc fingers in the N-terminal region
meristems, shoot meristems of the CESA proteins19. Membrane complexes probably
and flower meristems fit this
description.
Building the primary cell wall contain other proteins that aid microfibril formation
The growing cell wall has a fibreglass-like structure1,2 and that link the complexes to nearby microtubules for
VACUOLE (FIGS 1d,2),with crystalline CELLULOSE MICROFIBRILS that guidance along the membrane20.
A membrane-bound cellular are embedded in a matrix of complex polysaccharides, Each cellulose microfibril is formed from the
compartment, usually filled
which are divided into two classes BOX 1. PECTINS are spontaneous ‘bundling’ and crystallization of dozens
with a dilute watery solution.
Mature plant cells often have wall polysaccharides that are solubilized by aqueous of (1,4)-linked β-d-glucan chains, each made by a
very large central vacuoles. buffers and dilute acidic solutions or calcium chelators. CESA protein (FIG. 3c). Microfibrils are 3–5 nm wide
HEMICELLULOSES, on the other hand, require strong alkali and many micrometers in length — long enough to
WALL LOOSENING
for solubilization. Hemicelluloses are cellulose-binding wind around the circumference of a cell many times.
Modification of the cell wall
that enables it to extend in
polysaccharides, which together with cellulose form a It is possible that hemicelluloses, such as xyloglucan,
response to the wall stress that network that is strong yet resilient. Pectins are perhaps become trapped in the microfibril as it forms, resulting
is generated by cell turgor. the most complex polysaccharides in the living world in disordered regions21.

NATURE REVIEWS | MOLECULAR CELL BIOLOGY VOLUME 6 | NOVEMBER 2005 | 851

REVIEWS

Cytoplasm Two recent studies29,30 analysed patterns of gene

co-expression to identify a number of genes that are
Golgi apparatus
Vesicles associated with cellulose synthesis of secondary walls.
Many of these genes were not identified in previous
genetic screens for defective secondary walls, yet analy-
sis of lines bearing mutations in these genes revealed
Plasma membrane clear cell-wall defects. Further study of these genes
Cellulose synthase should broaden our understanding of the mechanisms
complex that are involved in cellulose synthesis.

Synthesis of matrix polysaccharides. In contrast to

cellulose, matrix polysaccharides possess a more
diverse set of glycosidic linkages and sugar residues
BOX 1 . Identification of the specific genes that
encode the relevant glycosyltransferases has proved
difficult, but almost 20 of these genes have recently
Cellulose Main pectin domains
Rhamnogalacturonan I
been characterized16.
Main hemicelluloses The CESA family belongs to a larger superfamily
Xyloglucan Homogalacturonan
of genes called CELLULOSE SYNTHASE-LIKE (CSL;
Arabinoxylan Xylogalacturonan see Online links box), which includes eight other
Arabinan gene families31, named CSLA, CSLB and so on, up to
Rhamnogalacturonan II
CSLH. CSL proteins contain sequence motifs that are
characteristic of β-glycosyltransferases, but lack the
Figure 2 | Structure of the primary cell wall. Cellulose microfibrils (purple rods) are
N-terminal region containing the zinc-finger domains
synthesized by large hexameric complexes in the plasma membrane, whereas hemicelluloses — which is found in CESA — that function in pro-
and pectins, which compose the matrix polysaccharides, are synthesized in the Golgi tein dimerization. Because of their similarity to, and
apparatus and are deposited to the wall surface by vesicles. For clarity, the hemicellulose– divergence from, CESA, CSL proteins are considered
cellulose network is shown on the left part of the cell wall without pectins, which are good candidates for the synthases that are localized
emphasized on the right part of the figure. In most plant species the main hemicellulose is in the Golgi and that form the β-DGLYCAN backbone of
xyloglucan (blue), while hemicelluloses such as arabinoxylans (grey) and mannans (not shown)
are found in lesser amounts. The main pectin polysaccharides include rhamnogalacturonan I
hemicelluloses such as xyloglucan, xylan, mannan and
and homogalacturonan, with smaller amounts of xylogalacturonan, arabinan, arabinogalactan I other β-d-glycans in the cell wall.
(not shown) and rhamnogalacturonan II. Pectin domains are believed to be covalently linked Recent studies indicate that CSLA genes encode
together 4 and to bind to xyloglucan by covalent and non-covalent bonds127,131. Neutral pectin β-mannan synthases, the enzymes that are required
polysaccharides (green) are also able to bind to cellulose surfaces121. See BOX 1 for further for the formation of the mannan backbone of certain
information on the structure of these polysaccharides. hemicelluloses. Dhugga et al.32 showed that galacto-
mannan of guar gum is synthesized by a CSLA gene
product. Galactomannan accumulates as a storage
If CESA enzymes add glucose residues to an exist- polysaccharide in guar-bean seeds and is used com-
ing glucan, how does the glucan chain initiate? Recent mercially as a food thickener. It has a β-mannan
studies indicate that a STEROL GLUCOSIDE might function backbone with galactose sidechains. Using a different
as the initial acceptor for chain elongation. Sterols are approach, Liepman et al.33 employed heterologous
common lipid components of plant cell membranes and expression of A. thaliana CSLA genes in insect cells
sterol-β-glucosides are commonly synthesized in plant to show that these genes encode mannan synthases,
plasma membranes, where CESA can use a sterol-glu- and they proposed that CSLA proteins synthesize the
PRIMARY CELL WALL
coside and uridine 5′-diphosphate-glucose to form short mannans and glucomannans of the growing cell wall.
The flexible extracellular matrix
that is deposited while the cell is sterol-linked glucans22. This idea is further supported by Heterologous expression of other CSL family members
expanding. studies showing that defects in sterol biosynthesis result in insect cells (which do not have a cell wall) could
in cell-wall gaps and reductions in cellulose content23. help to identify the function of these unique plant
SECONDARY CELL WALL
A membrane-bound endoglucanase (named proteins.
The flexible extracellular matrix
that is deposited while the cell is
KORRIGAN, or KOR) is also required for normal cel- In addition to the synthases that assemble the back-
still expanding is known as the lulose formation24,25, but its exact function is unclear. bones of the various matrix polysaccharides, other
primary cell wall. When kor mutants have defects in cytokinesis and cell elon- glycosyltransferases are required to add branches
expansion ceases, a secondary gation. Although they make (1,4)-linked β-d-glucan, to these glycans. Genes that encode some of these
wall is sometimes laid down
it is not properly crystallized into a microfibril. It has enzymes were recently identified (for a review, see
inside the primary wall, making
it stronger. been postulated that the KOR endoglucanase trims REF. 16). These important advances pave the way for
sterol residues from nascent glucan primers or trims rapid progress in defining the enzymes that synthesize
CELLULOSE MICROFIBRIL out-of-register glucans to aid crystallization of the wall polysaccharides.
A tough, inelastic fibre wrapped microfibril22. However, the protein is not specifically
in layers (lamellae) within the
plant cell wall. Composed of
associated with CESA complexes26,27 and kor mutants Forming a network. After the matrix polysaccharides
(1,4)-linked β-d-glucosyl have normal amounts of sterol glucoside28, so its action are secreted into the wall, they become associated
residues. on microfibril formation might be indirect. with newly synthesized cellulose microfibrils, as well

852 | NOVEMBER 2005 | VOLUME 6 www.nature.com/reviews/molcellbio

 REVIEWS

Box 1 | Structure of the main polysaccharides that are present in the growing cell wall
Cellulose. The primary structure of cellulose is an unbranched (1,4)-linked β-d-glucan. Many parallel glucans snap
into register to form a crystalline microfibril that is mechanically strong and highly resistant to enzymatic attack —
an almost ideal scaffold material. These long, crystalline ribbons are 3–5 nm wide and, in growing cells, are aligned
with each other, giving a structural bias to the cell wall (FIG. 1d).
Hemicelluloses. The backbone of hemicelluloses resembles that of cellulose. Hemicelluloses bind to cellulose, but
branches and other modifications in their structure prevent them from forming microfibrils by themselves.
Xyloglucan and arabinoxylan are two of the most abundant hemicelluloses. Details of their structure vary slightly
PECTINS
Group of complex among plant species. Xyloglucan has a backbone that is similar to that of cellulose, but it is decorated with xylose
polysaccharides that are branches on 3 out of 4 glucose residues. The xylose can also be serially appended with galactose (Gal) and fucose
extracted from the cell wall by (Fuc) residues. Arabinoxylan consists of a (1,4)-linked β-d-xylan backbone decorated with arabinose branches.
hot water, dilute acid or calcium Other residues, such as glucuronic acid and ferulic acid esters (FAE), are also attached in arabinoxylans that are
chelators. They include particularly abundant in cereal grasses. Mannans are also found in primary cell walls and probably function in the
homogalacturonan,
same way as xyloglucan and arabinoxylan.
rhamnogalacturonans I and II,
galactans, arabinans and other Pectins. This complex and heterogeneous group of polysaccharides consists of distinctive domains, which are
polysaccharides.
believed to be covalently linked together3,4,119. Rhamnogalacturan I consists of alternating residues of galacturonic
HEMICELLULOSES
acid and rhamnose, and probably has side branches that contain other pectin domains4. Homogalacturonan
Group of complex comprises a linear chain of galacturonic acid residues, whereas xylogalacturonan is modified by the addition of
polysaccharides, including xylose branches. The carboxyl groups of homogalacturonan and xylogalacturonan are often methyl esterified, a
xyloglucans, xylans and modification that ‘blocks’ the acidic group and reduces their ability to form gels. Rhamnogalacturonan II is a
mannans, that are extracted from complex pectin domain that contains 11 different sugar residues (for details see REFS 1,2) and forms dimers through
plant cell walls by use of strong
borate (B) esters. The neutral arabinans and arabinogalactans are also linked to the acidic pectins and it has been
alkali; characteristically they bind
tightly to the surface of cellulose
proposed that they promote wall flexibility120 and that they bind to the surface of cellulose121.
and have a backbone made up of Pectins and hemicelluloses are modified in many ways. See reviews1–4,119 for details. The structures of the molecules
(1,4)-β-d-glycans that resembles
cellulose.
described above are shown in Supplementary information S1 (figure).

POROSITY
Property that indicates how as with the pre-existing wall polymers, to form a a glycosidic bond with the free end of another xyloglu-
readily gases, liquids and other network that is both strong and extensible. Network can chain (FIG. 4). XET is a member of a large family of
materials can penetrate an object. formation involves spontaneous physico-chemical plant enzymes called XTH (see Online links box), for
MIDDLE LAMELLA
interactions between the wall polysaccharides and xyloglucan endotransglucosylase/hydrolase, because
The thin layer that connects two perhaps enzymatic crosslinking BOXES 2,3. At this level some members have hydrolase activity. XTHs and the
plant cells and is rich in pectin. of polymer–polymer interaction, the precise structure genes that encode them have been characterized in
of the cell wall has not yet been resolved in detail and detail42–46. DOUBLELABELLING EXPERIMENTS47 demonstrate
MATRIX POLYSACCHARIDES
remains a key issue for understanding how the cell wall that XET grafts new xyloglucans onto chains that are
Complex polysaccharides found
in the space between cellulose expands. already part of the wall network. The wall-strength-
microfibrils. They are Xyloglucan is an abundant hemicellulose in primary ening action of XET is supported by experiments in
traditionally divided into pectins cell walls, and is believed to crosslink microfibrils, which xyloglucans were added to excised pea stems,
and hemicelluloses. forming either a direct tether between microfibrils1,21,34 which consequently became stiffer41.
TURGOR PRESSURE
or an indirect link35 BOX 2. Breakdown of xyloglucan is Molecular modelling data indicated that some
Force generated by water increased after treatment with auxin, which is a potent members of the XTH family might target arabinoxy-
pushing outward on the plasma growth hormone36,37. The potential crosslinking role of lan and (1,3;1,4)-β-d-glucan48, hemicelluloses that
membrane and plant cell wall, xyloglucan was supported by experiments showing that are notably abundant in grass cell walls. However,
that results in plant rigidity. The
isolated walls were induced to creep upon digestion this hypothesis has yet to be confirmed experi-
loss of turgor pressure causes
wilting. with a fungal endoglucanase that digests xyloglucan38. mentally. An endotransglycosylase with specificity
Moreover, expression of a fungal xyloglucanase in pop- for mannans has recently been characterized 49 .
STEROL GLUCOSIDE lar gave transformants with longer and thicker stems39. Endotransglycosylases for xylans and other matrix
A molecule consisting of a A physical model of cell-wall growth, which was based polysaccharides presumably exist but have not yet
sterol that is linked to glucose
through a glycosidic bond.
on the thermodynamics of hydrogen-bonded networks, been discovered.
was proposed40. This model specifically predicts how In summary, cellulose microfibrils are linked
βDGLYCAN the abundance and size of xyloglucan ‘tethers’ should together by non-covalent interactions with matrix
A polymer built up of sugar affect wall enlargement. Experiments that attempted polysaccharides, which determine most of the physi-
residues connected by
to manipulate xyloglucan size in growing cells41 have cal properties of the cell wall. The complexity of the
glycosidic bonds; β-d- identifies
the particular stereochemical confirmed the predictions of this physical model. cell-wall network allows for many potential sites where
configuration of the sugar. Integration of newly secreted matrix polysaccha- loosening and expansion might be initiated, yet physics
rides into the existing network might also be mediated makes this a tricky business, as discussed below.
ENDOTRANSGLYCOSYLASES
by enzymes such as ENDOTRANSGLYCOSYLASES, which cut
Enzymes that cut a glycan and
ligate one of the fragments to
and ligate glycans together42–44. One such enzyme, The physical dilemma of wall enlargement
the free end of another polymer, called xyloglucan endotransglucosylase (XET), spe- The pliant walls of growing cells are under tremen-
usually of the same type. cifically cuts the xyloglucan backbone and re-forms dous tension, equivalent to 100–1000 atmospheres

NATURE REVIEWS | MOLECULAR CELL BIOLOGY VOLUME 6 | NOVEMBER 2005 | 853

REVIEWS

a b most cells. Fast growth responses — occuring in a few

seconds51 or even less than a second52 — are based on
rapid changes in wall extensibility without significant
alterations in wall composition or structure. By con-
trast, slower changes, such as the gradual decline in
growth that occurs as cells mature and as their walls
stiffen, involve substantial changes in wall composition
and crosslinking.
‘Wall loosening’ refers to a molecular modification
of the wall network that results in relaxation of wall
Rosette subunit stress. WALL STRESS RELAXATION might result from scission
of a stress-bearing crosslink or from sliding of such a
c (1,4)-β-D-glucan chain Cellulose microfibril crosslink along a scaffold; in both cases it results in a
reduction in wall stress without a substantial change
in wall dimensions53. Actual enlargement of the wall
occurs secondarily, as a consequence of cellular water
uptake, in response to the turgor relaxation that inevi-
tably accompanies wall stress relaxation.
Recently, four molecular mechanisms of wall loos-
ening have attracted recent attention and are reviewed
below. The candidate wall loosening agents include:
expansin, xyloglucan endotransglycolase/hydrolase,
endo-(1,4)-β-d-glucanase and HYDROXYL RADICALS. For
this discussion, it is useful to distinguish between two
classes of wall loosening agents, which I call ‘primary’
CESA Rosette subunit Rosette and ‘secondary’ wall loosening agents54. Primary agents
Figure 3 | The cellulose-synthesizing machinery of the cell. a | Immunogold labelling catalyse stress relaxation directly — this can be dem-
shows that CESA is localized to hexameric ‘particle rosettes’ in the plasma membrane6. The onstrated by applying the agent to isolated cell walls and
black circles represent gold nanoparticles that are attached to antibody against CESA. The
measuring either wall stress relaxation or wall extension
smallest subunit in the particle rosette is believed to be made of six CESA proteins. Particle
rosettes are sometimes found attached to cellulose microfibrils. Scale bar, 30 nm. b | This
(creep). By contrast, secondary agents modify the wall
model of a hexameric particle rosette shows how three different CESA proteins (shown in three without causing stress relaxation and they cause wall
different colours: α, orange; β, brown; γ, green) might be organized into rosette subunits and extension indirectly, by amplifying the physical effects of
then into a hexameric synthase complex7. CESA assembly into rosette subunits is thought to primary agents. Both types of loosening agent might be
involve oxidatively reversible disulfide bond formation between cysteines in the N-terminal zinc- important for the control of wall enlargement by plants.
finger region of CESA19. c | A model of how CESA complexes synthesize a cellulose
microfibril7. Each CESA protein can synthesize a single (1,4)-linked β-D-glucan chain. Cellulose
Expansin
is formed as a crystalline ribbon that is composed of many such glucans. In this model, 36 β-D-
glucan chains are formed by a particle rosette, which is composed of a hexamer of CESA Growing plant cells characteristically exhibit ACID
55–57
hexamers. Part a of the figure is reproduced, with permission, from REF.6 © American Society GROWTH . Isolated cell walls also exhibit this phenom-
of Plant Biologists (1999). Parts b and c of the figure are modified, with permission, from REF.19 enon, which results from the action of pH-dependent
© Oxford University Press (2002). wall-loosening proteins named EXPANSINS58. The pH
of the cell wall of growing cells is typically between
4.5 and 6, which is the range in which acidification
of tensile stress. Wall stress results from the stretch- activates expansin activity. This is biologically impor-
ing of wall polymers as they resist cell turgor pres- tant because a number of agents alter cell growth, at
sure and provides the mechanical energy that is least in part, by inducing the cell to alter its wall pH
required to extend the cell wall. But how can this — through control of the plasma membrane proton
DOUBLELABELLING thin material irreversibly extend its surface area pump (H+-ATPase). The plant growth responses that
EXPERIMENTS without risking a fatal blowout? It does so by a involve changes in cell-wall pH include stem and root
Experiments in which two
mechanism of carefully controlled polymer creep, tropisms (bending toward or away from a stimulus);
different tags (such as the
radioisotopes 3H and 14C) are in which the matrix yields, allowing the cellulose hormone-induced growth; light-induced stimulation
incorporated in and attached to a microfibrils to move apart50. Although the mechani- of leaf expansion and inhibition of stem elongation;
molecule. Useful for tracing the cal force for this polymer motion comes from wall responses of shoots, roots and leaves to water deficits
origin and fate of a molecule that stress (and therefore turgor pressure), control of and salt stress; initiation and early outgrowth of root
undergoes complicated
processing.
the process resides in the selective loosening and hairs; and rapid growth in response to the fungal toxin
shifting of load-bearing linkages between cellulose fusicoccin.
WALL STRESS RELAXATION microfibrils.
Reduction in mechanical stress In the longer term, wall expansion must be matched A role for expansins in cell growth. Four lines of evi-
in the cell-wall network,
by the synthesis and integration of new wall materials dence support the idea that expansins (see Online links
because of slippage or scission
of load-bearing polymers in the to prevent thinning to the point of mechanical insta- box) regulate cell wall enlargement in growing cells.
cell wall; wall loosening bility. However, expansion and synthesis of the wall First, when the acid-growth behaviour of isolated cell
stimulates wall stress relaxation. are separate processes that are only loosely coupled in walls is eliminated by heat or protease treatment, it

854 | NOVEMBER 2005 | VOLUME 6 www.nature.com/reviews/molcellbio

 REVIEWS

Box 2 | How do hemicelluloses form a network with cellulose microfibrils?

At least five ideas (see figure, Cellulose
a–e) have been considered
and involve direct or indirect
linkages between cellulose a b
Arabinoxylan
microfibrils. Hemicelluloses
(grey and blue strands) may Xyloglucan c
e
form a primary network with Arabinan
cellulose, and may also be A-F-F-A
linked to acidic pectins (red d
strands). Additionally, neutral
pectin polysaccharides, such
as arabinans (green strands),
are able to bind the cellulose
surfaces as well121.
Pectin polysaccharides
• Hemicelluloses might
spontaneously bind to the
surfaces of cellulose microfibrils and tether adjacent microfibrils together21,34 (see figure, a).
• Xyloglucan might become entrapped during formation of the ordered microfibril122,123 (see figure, b). The
untrapped remainder of the xyloglucan would be free to bind to other cellulose surfaces or to other matrix
polymers, thereby anchoring the microfibril firmly to its neighbours. This hypothesis might explain why a
significant xyloglucan fraction is released from the microfibril only after treatments that cause microfibril swelling
(for example, treatment with a strong alkali). It could also explain why a cellulose-specific endoglucanase stimulates
cell enlargement: digestion of the noncrystalline region of a cellulose molecule results in the release of the trapped
xyloglucan and the microfibril is freed from its tether124.
• Cellulose microfibrils might be simply coated with xyloglucans (blue strands), which adhere to other matrix
polysaccharides, without direct linkage between microfibrils (see figure, c). This idea is based on studies showing
that the main polysaccharides can be selectively extracted from the growing wall without evidence for covalent
linkage with other polysaccharide species35.
• Xyloglucans (blue strands) might be covalently attached to pectin polysaccharides (red strands), forming a
macromolecule that anchors the microfibrils by sticking of xyloglucan to cellulose surfaces (see figure, d). This is a
variation of the earliest molecular model of the plant cell wall125, which was subsequently dismissed owing to lack of
evidence of the covalent linkage between xyloglucan and pectin. However, recent studies126,127 report that a
proportion of xyloglucan is covalently attached to acidic residues, possibly homogalacturonan. The nature of the
hypothetical xyloglucan–pectin linkage has not been established.
• Arabinoxylans (grey strands) might bind cellulose and be crosslinked by ferulic acid esters (A-F-F-A) (see figure, e).
This type of phenolic crosslink might also crosslink other hemicelluloses and pectins, particularly in gross cell walls.

can be almost fully restored by the addition of purified walls begin to creep at a fast rate58. By contrast, walls
expansin proteins58. This indicates that expansins are that were treated with a fungal endoglucanase, which
HYDROXYL RADICAL primary wall loosening agents and that their action hydrolyses xyloglucan, required prolonged digestion
The most active form of reactive alone is sufficient to restore extensibility to cell walls. before the walls began to extend38. Whereas endoglu-
oxygen species, consisting of a Second, addition of exogenous expansin to growing canase digestion physically weakened the cell wall (as
free hydroxyl group in which
cells stimulates their growth59–61. These results show measured by plastic and elastic compliances), expansin
oxygen is missing an electron in
its outermost shell. It is a strong that expansin can stimulate cell enlargement under the treatment did not. These compliances are measured
oxidant that can steal an normal operating conditions of the cell wall and that in stress/strain experiments and changes indicate
electron from — and thereby endogenous expansins are at least partially limiting modification of cell-wall structure and the degree of
damage — polysaccharides, for cell growth. Third, ectopic expression of expansin crosslinking. On the other hand, expansins promote
proteins, lipids, nucleic acids
and other classes of organic
genes stimulates plant growth, whereas suppression of wall stress relaxation, but endoglucanases have not
molecules. expansins by gene silencing decreases plant growth62–65, been found to have this ability. Similarly, expansins and
indicating that altered expression of expansin genes can endoglucanases have distinctive mechanical effects on
ACID GROWTH modify plant growth. Last, endogenous expansin-gene artificial cellulose–xyloglucan composites73. It is there-
Faster cell elongation under
expression correlates with the onset, increase and ces- fore clear that expansins and endoglucanases modify
acidic conditions.
sation of cell growth66–72. Expansin genes are expressed walls in distinctive ways and with different kinetics.
EXPANSINS at the right time and the right place to exert control of Sequence analyses74,75 indicate that expansins con-
Wall loosening proteins that cell growth. sist of two domains: an N-terminal domain (~15 kDa)
induce wall stress relaxation with distant sequence similarity to the catalytic domain
and irreversible wall extension
in a pH-dependent manner, but
Exploring expansin function. How expansins catalyse of the family-45 endoglucanases; and a C-terminal
they do not hydrolyze wall wall enlargement remains unclear. They function domain (~10 kDa) that is related to a family of grass-
polymers. rapidly — within seconds of expansin addition, isolated pollen allergens of unknown function. Despite the

NATURE REVIEWS | MOLECULAR CELL BIOLOGY VOLUME 6 | NOVEMBER 2005 | 855

REVIEWS

Box 3 | Formation of pectin networks by covalent and ionic bonds

The distinctive pectin domains are believed to be a c
covalently crosslinked to each other (see figure, part Ca2+ Ca2+ Ca2+

a), but the nature of the crosslink has not been Ca2+ Ca2+

determined. In addition, two other types of linkage Ca2+

COO–
involving boron (part b) and calcium (part c) are B– O COOMe O–
important crosslinking mechanisms. O
Ca2+
• Part a of the figure shows a model of how the pectin Ca2+ O
Ca2+ Ca2+ Ca2+ MeOOC O
domains may be covalently linked together to form a O
–
OOC
massively large macromolecular pectin network.
This is a simplified version of a recent model by
COO–
Vincken et al.4, in which rhamnogalacturonan I O COO–
O
O
serves as the backbone and the other pectin domains Ca2+
are attached as branches. Homogalacturonans are O
O –
OOC O
ionically crosslinked by calcium (part c) whereas –
OOC
boron crosslinked rhamnogalacturonan II through
diester linkages (part b). O COO–
COO–
O
• Rhamnogalacturonan II forms dimers through a b O
Ca2+
borate ester bond (part b). This crosslinking is O– O
O –OOC
important for normal wall formation as well as for –
OOC
O
the control of wall porosity and wall thickness. O COO–
O COO–
–
OOC O Sidechain
– O
• Homogalacturonan (also known as polygalacturonic OOC
O
COO–
2+ O O COOMe O
acid) forms stiff gels through Ca -mediated O
O
O O
crosslinking of its carboxyl groups through ionic O Api B– Api O Ca2+
O O
O
and COORDINATE BONDS (part c). Growing cells usually O O
O MeOOC O
O–
synthesize homogalacturonan in which ~75% of the O Sidechain O COO–
COO – –
OOC
–OOC
carboxyl groups (COO–) are methyl esterified –
OOC
O O
(COOMe) — this modification removes the negative
charge of the carboxylate ion and blocks its ability to O–
undergo Ca2+ crosslinking. Highly esterified
homogalacturonans do not form stiff gels and their
secretion might help the expanding wall to remain
pliant. Carboxyl-based crosslinking sites are
unmasked later, as the cells cease growth, owing to Rhamnogalacturonan I B– Borate diester
the action of pectin methylesterases. These Homogalacturonan Ca2+ Calcium
methylesterases, which are secreted by plant cells
Xylogalacturonan Arabinan
into their wall space, hydrolyse the methylesters and
free the carboxyl group for Ca2+ crosslink formation Rhamnogalacturonan II
and gel formation.

similarity with endoglucanases, no enzymatic activity a hydrolytic enzyme. As soon as expansin enters the
has been found that accounts for the action of expansin wall it stimulates extension, and subsequent removal
on the wall76–78. A report79 that β-expansin exhibited of expansin restores the wall to an inextensible state76.
protease activity when it was recombinantly expressed This indicates that expansin does not alter gross wall
in yeast was later refuted78, with the activity apparently structure or the degree of crosslinking. Last, 2 M urea,
being the result of an induced yeast protease80. which disrupts hydrogen bonding, mimics some of
Expansins are believed to disrupt non-covalent the physical effects of expansin on the cell wall77. On
binding of wall polysaccharides to one another. This the basis of these and other results, we have proposed
concept has been supported by several studies. First, that expansin functions to dissociate a polysaccharide
expansin weakens paper, which is a hydrogen-bonded complex that links microfibrils together76,77.
network of cellulose fibrils, but this weakening action
does not involve cellulose hydrolysis 77 . Second, Expansins: growth and beyond? Genomic and phylo-
expansin synergistically enhances the hydrolysis of genetic analyses show that plant expansins consist of
crystalline cellulose by cellulases. Because glucan a large superfamily that is divided into four divergent
accessibility is the rate-limiting step in cellulase action, families81. It has been demonstrated that two families
this result could indicate that expansin promotes contain members that have the ability to extend walls
the release of glucans on the surface of the cellulose (EXPA or α-expansin, and EXPB or β-expansin),
COORDINATE BOND
Chemical bond involving the
microfibril, making them available for enzymatic whereas the functions of the other two related families
sharing of a pair of electrons, attack. Third, expansin does not gradually and progres- (EXLA and EXLB, for expansin-like family A and B)
each supplied by one atom. sively weaken the cell wall, as would be expected for remain to be established. A. thaliana has 36 members

856 | NOVEMBER 2005 | VOLUME 6 www.nature.com/reviews/molcellbio

 REVIEWS

a wall creep have yielded only negative results (J. Rose

Xyloglucans and D.C., unpublished data; REFS 94,95). These data
contrast with results obtained using artificial cellu-
lose–xyloglucan composites, in which XTH induced
creep in BIAXIAL STRAIN ASSAYS73. Such results indicate that
XTH can loosen some wall-like networks, but they also
point out that results that were obtained with these
XTH artificial composites cannot always be extrapolated to
the behaviour of real cell walls. When isolated walls of
b azuki bean HYPOCOTYLS were incubated with XTH, they
became more extensible in force/extension assays96
(this is not the same as creep53). This is apparently the
only published report showing that XTH reduces the
mechanical strength of isolated walls. However, the
XTH long incubation times that were used in these experi-
ments (48 hours at 37oC) indicate very weak enzyme
c activity. Low-level contamination with other enzymes
is a potential concern in such situations. Also, the
walls were pre-treated with boiling methanol, which
— although it kills most enzyme activity — does
not reduce expansin activity56,58. Therefore, the cell
walls that were used in these experiments probably
XTH
contained active endogenous expansin, as well as
exogenous XTH.
However, another study that used living tissues (not
Figure 4 | The activity of xyloglucan endotransglucosy- isolated walls) demonstrated that xyloglucan metabo-
lase/hydrolase (XTH) as an endotransglucosylase. The
lism by endogenous XTH is a potential control point for
enzyme carries out two reactions, first a scission of a
glycosidic bond in the xyloglucan backbone, followed by the cell elongation41. Addition of long xyloglucans, which
re-formation of the bond with a second xyloglucan chain. have the potential to crosslink cellulose microfibrils,
a | Two xyloglucan chains are shown (blue and green). XTH reduced growth of pea-stem segments by more than
binds to one of the chains. b | Upon cutting the xyloglucan 50%. The opposite response, a growth stimulation and
backbone, one of the strands remains covalently attached to substantial xyloglucan solubilization, was obtained when
the catalytic site of the enzyme. The free end of a second
short xyloglucan oligosaccharides (XGOs) were used.
xyloglucan moves into place for the second reaction. c | After
ligation, a hybrid xyloglucan is formed, which is indicated as a
When XGOs are used as acceptors in a transglucosyla-
xyloglucan of two colours and a xyloglucan fragment is tion reaction by XET, the result is similar to the action of
released. hydrolase — the xyloglucan chain length is reduced.
These experiments — in which xyloglucan is arti-
ficially manipulated — confirm the significance of
of this superfamily, whereas rice has 58 members. The xyloglucans for cell enlargement. However, the milli-
biological function of individual expansin genes is molar level of XGOs used in these experiments is much
currently being studied. Gene-expression studies have higher than that found in vivo97, so the case for XET
shown that many genes are related to the growth of as a primary wall-loosening enzyme in vivo cannot be
specific cell types, whereas others promote wall loos- considered strong on the basis of this result. Because
ening in other situations, such as fruit softening82,83, large xyloglucans are continually being deposited at
84 85
ABSCISSION and pollination . A challenge for the future the inner surface of the growing wall, a strengthening
will be to elucidate the specific roles of the genes that function for XET seems more probable.
belong in this superfamily.
Seeing XET in action. The activity of XET can be visu-
ABSCISSION Xyloglucan endotransglucosylase/hydrolase alized with the use of fluorescently labelled XGOs98,99.
The process by which old parts
of a plant break off naturally
It has been postulated that XTHs carry out vari- When roots were labelled in this way, XET activity was
(for example, leaves). ous functions46; these include wall loosening86, wall greatest in the elongation region100, as well as at sites of
strengthening87, integrating new xyloglucans into the root-hair initiation, where the thick outer epidermal
BIAXIAL STRAIN ASSAYS wall47, trimming xyloglucan strands that are not tightly wall begins to bulge out. Studies of XTH gene expres-
Procedures in which a material
stuck to the surface of cellulose88, fruit softening89, and sion101 indicate that four XTH genes might account
is stretched not in one direction,
but in two directions, as hydrolysing xyloglucans90,91 particularly during xylem for these XET activities. A recent study102 noted that
happens with the membrane of formation92,93. XGO fluorescence was associated with cellulose micro-
an expanding balloon. Despite assertions in the literature that XTH causes fibrils in elongating cells, whereas in dividing (non-
wall loosening, there is little direct evidence in support elongating) cells the fluorescence pattern was diffuse.
HYPOCOTYLS
The stem region of a seedling
of XTH as a primary wall-loosening agent – that is, The authors proposed that some XET isoforms select
below the cotyledons (seed an agent that induces wall stress relaxation and exten- xyloglucan that is tightly bound to cellulose, whereas
leaves). sion by itself. Attempts to measure XTH-induced cell others function less specifically. Another possibility is

NATURE REVIEWS | MOLECULAR CELL BIOLOGY VOLUME 6 | NOVEMBER 2005 | 857

REVIEWS

that newly secreted xyloglucans are particularly acces- In summary, the work described in the last two
sible to tagging by XET under the conditions of this sections indicates that xyloglucan–cellulose interac-
assay, and they soon become bound to nascent cellu- tions are important determinants of the mechanical
lose microfibrils, which are synthesized at high rates and growth properties of the cell wall. The role of the
in elongating cells. This gives rise to a fibrillar pattern XTH family of enzymes in the regulation of cell-wall
of fluorescence in such cells (see cellulose microfibrils loosening versus wall strengthening is unclear but
in FIG. 1d). the latter role has more support. The role of endo-
XTH gene expression is high in regions of active (1,4)-β-d-glucanases in wall loosening merits greater
wall formation — that is, in elongation zones and in attention.
regions where wall deposition continues after cell
enlargement has ceased or where other forms of wall Hydroxyl radical (•OH)
remodelling occur87,103–105. But what are the conse- •OH is a highly active form of reactive oxygen species,
quences of manipulating XTH levels in plants? Herbers which have important roles in signalling and cell death.
et al.106 suppressed XTH gene expression in tobacco Recent studies have advanced the new idea that •OH
plants using antisense methods; the enzyme activity is harnessed by growing cells to loosen their cell walls
was reduced as much as 56% and the xyloglucan size and stimulate cell enlargement112–116. This hypothesis
increased by 20%, but effects on plant growth were is based on the observation that •OH can cleave wall
not noted. An XTH mutant with a subtle phenotype polysaccharides (along with almost everything else it
in xylem development was recently characterized93; touches) by nonenzymatically removing a hydrogen
the A. thaliana xth27 mutant had fewer tertiary veins, atom from polysaccharides113. This idea is supported by
and where tertiary veins did form, the TRACHEARY ELE results showing that extension of isolated walls could
MENTS were shorter and misshapen. Although XTH27 be induced by artificially generated •OH114. Moreover,
is expressed in many cells in the plant, a developmental plant cells produced •OH in an auxin-dependent
phenotype was only observed in the tertiary veins of manner and free-radical quenchers suppressed auxin-
the mutant plants. These results indicate that XTH27 induced growth116.
is required for proper morphogenesis of tertiary veins, It has been postulated that endogenous •OH might
but the interpretation of this phenotype in terms of be produced nonenzymically by copper ions that
wall loosening, xyloglucan removal from secondary are bound to the cell wall112 or by wall peroxidases115
walls or other potential biochemical functions of XTH from the superoxide anion and hydrogen peroxide.
is less clear. Hydrogen peroxide is formed by monovalent O 2
reduction by a plasma-membrane NAD(P)H oxidase
Endo-(1,4)-β-D-glucanase and takes part in various defence responses, as well as
In addition to XTH, plants also have a family of secreted in hormone signalling.
endo-(1,4)-β-d-glucanases (sometimes called ‘cellu- Therefore, various observations implicate the
lases’). These enzymes belong to glycoside hydrolase involvement of •OH in auxin-induced cell growth.
family 9 (see Carbohydrate-active enzymes in Online However, in published studies the total amount of
links box), of which there are 25 family members in wall extension that was induced by •OH was small
A. thaliana. Three are membrane-bound endogluca- (~1% extension) and our attempts to induce more real-
nases (KOR and its PARALOGUES) that are involved in cel- istic amounts of extension by •OH treatments led to
lulose formation. The remaining proteins are secreted wall breakage rather than extension (L. Zhao and D.C.,
enzymes, mostly of unknown function. unpublished data). In comparison, ‘acid growth’ of
Although the expression of plant endoglucanase walls results in a 40% to 100% extension before the wall
genes has been studied for years in connection is weakened to the point of breakage56. Furthermore, it
with fruit softening, abscission and growth, the is doubtful that the amount of •OH that is produced by
enzymes’ biochemical properties, substrate specifi- growing cells approaches the high concentration that
cities and potential function for cell-wall loosening is required to cause cell-wall extension in vitro; such
have drawn surprisingly little experimental atten- concentrations might be expected to cause widespread
tion107,108. Potential wall substrates include cellulose damage to living cells.
and xyloglucan. Ohmiya et al.109,110 examined two If •OH does function as a bona fide wall-loosening
such enzymes from poplar (PopCel1 and PopCel2) agent in living cells, its production and release must be
TRACHEARY ELEMENTS and presented indirect evidence that they digest non- carefully controlled, so that its destructive reactivity
Specialized cells in the xylem of crystalline regions of cellulose. The authors proposed is targeted specifically to xyloglucan and other load-
vascular plants that are that the enzymes loosen walls by causing the release bearing polysaccharides. How this might be achieved
responsible for the conductance
of xyloglucans trapped in cellulose microfibrils. has yet to be determined, as •OH probably reacts with
of water as well as providing
mechanical support. Overexpression of PopCel1 in A. thaliana resulted in the first suitable molecule it encounters (including pro-
enhanced plant growth and increased plastic exten- teins, lipids and other biomolecules). Evidence of •OH-
PARALOGUES sibility of the cell walls. Conversely, antisense sup- induced polysaccharide splitting in growing cell walls
Genes or gene families that pression of these enzymes in poplar led to reduced might be obtained by finding the predicted products
originated from a common
ancestral sequence by a
leaf growth, by as much as 32%. Similar results were of •OH action in growing cell walls117. Evidence for
duplication event, not involving obtained in A. thaliana by silencing a related endog- such products has been found in softening fruit118, but
speciation. lucanase gene111. not in growing cells112.

858 | NOVEMBER 2005 | VOLUME 6 www.nature.com/reviews/molcellbio

 REVIEWS

Conclusions increased wall extensibility and cell growth. •OH

The growing cell wall in plants is a thin, strong and has been proposed as a wall loosening agent that
pliant extracellular layer that is composed of cellulose nonenzymatically cuts wall polysaccharides, but
microfibrils. These microfibrils are embedded in a questions remain about the physiological signifi-
hydrated matrix that is made of complex polysac- cance of this loosening mechanism.
charides and a small amount of structural proteins. Genetics and genomics have proved to be key tools
Microfibrils are synthesized by membrane complexes in the past decade for uncovering the enzymes that
that contain cellulose synthases, which are encoded assemble sugars into the diverse polysaccharides of the
by three CESA genes. Matrix polysaccharides are syn- plant cell wall and for characterizing the machinery
thesized in the Golgi apparatus and become integrated that underlies the formation and growth of the wall.
into the existing wall by enzymes and by spontaneous Enzymes that synthesize and modify the wall are
binding mechanisms. Recent studies have identified members of moderate to large multigene families, so
the genes that encode the glycosyltransferases that a challenge for the future will be to dissect the specific
synthesize some of the glycosidic linkages in xyloglu- functions of all of these genes. We have only recently
cans, mannans and pectins. begun to characterize the transferases that form spe-
Cell-wall enlargement begins with wall stress cific linkages in the cell wall (there are over 400 genes
relaxation, which allows the cells to take up water for such transferases in A. thaliana16), but we can
and physically enlarge. Expansins are a group of anticipate rapid progress in the coming years. Further
nonenzymatic wall proteins that induce wall stress progress on more recalcitrant problems, such as under-
relaxation and extension. They mediate ‘acid- standing the assembly of the macromolecular wall and
induced growth’ by disrupting the non-covalent the molecular bases of crosslinking, will be made by
linkages that hold microfibrils in place in the cell combining these approaches with new biochemical
wall. Xyloglucan endotransglucosylase cuts and and biophysical methods for assessing wall structure
joins xyloglucans. The consequences of this bio- and the changes that underlie its growth. This poses
chemical activity for wall properties are still being daunting technical challenges, but should reveal key
investigated. Plant endo-(1,4)-β-d-glucanases might missing pieces in the puzzle of cell-wall growth, and
digest the noncrystalline regions of cellulose micro- perhaps bring Robert Frost’s poetic appreciation of the
fibrils and release trapped xyloglucans, resulting in plant form into new light.

1. Carpita, N. C. & Gibeaut, D. M. Structural models of 12. Arioli, T. et al. Molecular analysis of cellulose biosynthesis 25. Nicol, F. et al. A plasma membrane-bound putative
primary cell walls in flowering plants: consistency of in Arabidopsis. Science 279, 717–720 (1998). endo-1,4-β-D-glucanase is required for normal wall
molecular structure with the physical properties of the Genetic evidence identifying cellulose synthase genes. assembly and cell elongation in Arabidopsis. EMBO J. 17,
walls during growth. Plant J. 3, 1–30 (1993). 13. Taylor, N. G., Howells, R. M., Huttly, A. K., Vickers, K. & 5563–5576 (1998).
A detailed technical review of the structure of cell Turner, S. R. Interactions among three distinct CesA 26. Zuo, J. et al. KORRIGAN, an Arabidopsis endo-1,4-β-
wall polysaccharides and how they might be proteins essential for cellulose synthesis. Proc. Natl Acad. glucanase, localizes to the cell plate by polarized targeting
arranged in the wall. Sci. USA 100, 1450–1455 (2003). and is essential for cytokinesis. Plant Cell 12, 1137–1152
2. O’Neill, M. A. & York, W. S. The Plant Cell Wall 14. Burn, J. E., Hocart, C. H., Birch, R. J., Cork, A. C. & (2000).
(ed. Rose, J. K. C.) 1–54 (Blackwell, Oxford, 2003). Williamson, R. E. Functional analysis of the cellulose 27. Szyjanowicz, P. M. et al. The irregular xylem 2 mutant is an
3. Willats, W. G., McCartney, L., Mackie, W. & Knox, J. P. synthase genes CesA1, CesA2, and CesA3 in allele of korrigan that affects the secondary cell wall of
Pectin: cell biology and prospects for functional analysis. Arabidopsis. Plant Physiol. 129, 797–807 (2002). Arabidopsis thaliana. Plant J. 37, 730–740 (2004).
Plant Mol. Biol. 47, 9–27 (2001). 15. Burton, R. A., Shirley, N. J., King, B. J., Harvey, A. J. & 28. Robert, S., Mouille, G. & Hofte, H. The mechanism and
4. Vincken, J. P. et al. If homogalacturonan were a side chain Fincher, G. B. The CesA gene family of barley. Quantitative regulation of cellulose synthesis in primary walls: lessons
of rhamnogalacturonan I. Implications for cell wall analysis of transcripts reveals two groups of co-expressed from cellulose-deficient Arabidopsis mutants. Cellulose 11,
architecture. Plant Physiol. 132, 1781–1789 (2003). genes. Plant Physiol. 134, 224–236 (2004). 351–364 (2004).
Discusses alternative models for how pectin 16. Scheible, W. R. & Pauly, M. Glycosyltransferases and cell 29. Persson, S., Wei, H., Milne, J., Page, G. P. & Somerville, C. R.
domains may be connected to one another. wall biosynthesis: novel players and insights. Curr. Opin. Identification of genes required for cellulose synthesis by
5. Iwai, H., Masaoka, N., Ishii, T. & Satoh, S. A pectin Plant Biol. 7, 285–295 (2004). regression analysis of public microarray data sets. Proc. Natl
glucuronyltransferase gene is essential for intercellular 17. Fagard, M. et al. PROCUSTE1 encodes a cellulose Acad. Sci. USA 102, 8633–8638 (2005).
attachment in the plant meristem. Proc. Natl Acad. Sci. synthase required for normal cell elongation specifically in 30. Brown, D. M., Zeef, L. A., Ellis, J., Goodacre, R. &
USA 99, 16319–16324 (2002). roots and dark-grown hypocotyls of Arabidopsis. Plant Turner, S. R. Identification of novel genes in Arabidopsis
6. Kimura, S. et al. Immunogold labeling of rosette terminal Cell 12, 2409–2424 (2000). involved in secondary cell wall formation using expression
cellulose-synthesizing complexes in the vascular plant 18. Taylor, N. G., Howells, R. M., Huttly, A. K., Vickers, K. & profiling and reverse genetics. Plant Cell 17, 2281–2285
Vigna angularis. Plant Cell 11, 2075–2085 (1999). Turner, S. R. Interactions among three distinct CesA (2005).
Showed that cellulose synthase is associated proteins essential for cellulose synthesis. Proc. Natl Acad. 31. Richmond, T. A. & Somerville, C. R. The cellulose synthase
with particle rosettes or terminal complexes in Sci. USA 100, 1450–1455 (2003). superfamily. Plant Physiol. 124, 495–498 (2000).
plants. 19. Kurek, I., Kawagoe, Y., Jacob-Wilk, D., Doblin, M. & 32. Dhugga, K. S. et al. Guar seed β-mannan synthase is a
7. Doblin, M. S., Kurek, I., Jacob-Wilk, D. & Delmer, D. P. Delmer, D. Dimerization of cotton fiber cellulose synthase member of the cellulose synthase super gene family.
Cellulose biosynthesis in plants: from genes to rosettes. catalytic subunits occurs via oxidation of the zinc-binding Science 303, 363–366 (2004).
Plant Cell Physiol. 43, 1407–1420 (2002). domains. Proc. Natl Acad. Sci. USA 99, 11109–11114 Elegant biochemical evidence that a member of the
8. Saxena, I. M. & Brown, R. M. Jr. Cellulose biosynthesis: (2002). CSLA gene family encodes the β-mannan synthase.
current views and evolving concepts. Ann. Bot. (Lond.) 96, 20. Lloyd, C. & Chan, J. Microtubules and the shape of plants 33. Liepman, A. H., Wilkerson, C. G. & Keegstra, K. Expression
9–21 (2005). to come. Nature Rev. Mol. Cell Biol. 5, 13–22 (2004). of cellulose synthase-like (Csl) genes in insect cells reveals
9. Ray, P. M. Radiographic study of cell wall deposition in 21. Hayashi, T. Xyloglucans in the primary cell wall. Ann. Rev. that CslA family members encode mannan synthases.
growing plant cells. J. Cell Biol. 35, 659–674 (1967). Plant Physiol. Plant Mol. Biol. 40, 139–168 (1989). Proc. Natl Acad. Sci. USA 102, 2221–2226 (2005).
10. Proseus, T. E. & Boyer, J. S. Turgor pressure moves 22. Peng, L., Kawagoe, Y., Hogan, P. & Delmer, D. Sitosterol- 34. Fry, S. C. Cellulases, hemicelluloses and auxin-stimulated
polysaccharides into growing cell walls of Chara corallina. β-glucoside as primer for cellulose synthesis in plants. growth: A possible relationship. Physiol. Plant. 75,
Ann. Bot. 95, 967–976 (2005). Science 295, 147–150 (2002). 532–536 (1989).
Showed that turgor pressure facilitates movement of 23. Schrick, K. et al. A link between sterol biosynthesis, the 35. Talbott, L. D. & Ray, P. M. Molecular size and separability
polysaccharides into the cell wall by multiple means. cell wall, and cellulose in Arabidopsis. Plant J. 38, features of pea cell wall polysaccharides. Implications for
11. Pear, J. R., Kawagoe, Y., Schreckengost, W. E., 227–243 (2004). models of primary wall structure. Plant Physiol. 92,
Delmer, D. P. & Stalker, D. M. Higher plants contain 24. Lane, D. R. et al. Temperature-sensitive alleles of RSW2 357–368 (1992).
homologs of the bacterial celA genes encoding the link the KORRIGAN endo-1,4-β-glucanase to cellulose Showed that matrix polysaccharides are associated
catalytic subunit of cellulose synthase. Proc. Natl Acad. synthesis and cytokinesis in Arabidopsis. Plant Physiol. with each other mostly non-covalently and
Sci. USA 93, 12637–12642 (1996). 126, 278–288 (2001). presented an alternative model of wall architecture.

NATURE REVIEWS | MOLECULAR CELL BIOLOGY VOLUME 6 | NOVEMBER 2005 | 859

REVIEWS

36. Labavitch, J. M. & Ray, P. M. Relationship between Purification of two related proteins, later named 78. Li, L. C. & Cosgrove, D. J. Grass group I pollen allergens
promotion of xyloglucan metabolism and induction of expansins, that cause pH-dependent wall extension. (β-expansins) lack proteinase activity and do not cause
elongation by IAA. Plant Physiol. 54, 499–502 (1974). 59. Link, B. M. & Cosgrove, D. J. Acid-growth response and wall loosening via proteolysis. Eur. J. Biochem. 268,
37. Nishitani, K. & Masuda, Y. Auxin-induced changes in the α-expansins in suspension cultures of bright yellow 2 4217–4226 (2001).
cell wall structure: changes in the sugar compositions, tobacco. Plant Physiol. 118, 907–916 (1998). 79. Grobe, K., Becker, W. M. & Petersen, A. Grass group I
intrinsic viscosity and molecular weight distributions of 60. Cosgrove, D. J. et al. The growing world of expansins. allergens (β-expansins) are novel, papain-related
matrix polysaccharides of the epicotyl cell wall of Vigna Plant Cell Physiol. 43, 1436–1444 (2002). proteinases. Eur. J. Biochem. 263, 33–40 (1999).
angularis. Physiol. Plant. 52, 482–494 (1981). 61. Fleming, A. J., McQueen-Mason, S., Mandel, T. & 80. Poppelmann, M., Becker, W. M. & Petersen, A.
38. Yuan, S., Wu, Y. & Cosgrove, D. J. A fungal Kuhlemeier, C. Induction of leaf primordia by the cell wall Combination of zymography and immunodetection to
endoglucanase with plant cell wall extension activity. Plant protein expansin. Science 276, 1415–1418 (1997). analyze proteins in complex culture supernatants.
Physiol. 127, 324–333 (2001). 62. Pien, S., Wyrzykowska, J., McQueen-Mason, S., Smart, Electrophoresis 23, 993–997 (2002).
First demonstration that plant cell walls can extend C. & Fleming, A. Local expression of expansin induces the 81. Kende, H. et al. Nomenclature for members of the
as a result of hydrolysis by an endoglucanase entire process of leaf development and modifies leaf expansin superfamily of genes and proteins. Plant Mol.
(fungal, not plant), but with characteristics very shape. Proc. Natl Acad. Sci. USA 98, 11812–11817 Biol. 55, 311–314 (2004).
different from expansin action. (2001). This article by the expansin research community
39. Park, Y. W. et al. Enhancement of growth and cellulose Used transient local micro-induction of EXPA genes defines expansins and recommends naming
accumulation by overexpression of xyloglucanase in on the shoot apical meristem and the flanks of leaf conventions.
poplar. FEBS Lett. 564, 183–187 (2004). primordia to demonstrate that EXPA overexpression 82. Rose, J. K., Lee, H. H. & Bennett, A. B. Expression of a
40. Veytsman, B. A. & Cosgrove, D. J. A model of cell wall can markedly stimulate plant cell growth. divergent expansin gene is fruit-specific and ripening-
expansion based on thermodynamics of polymer 63. Cho, H. T. & Cosgrove, D. J. Altered expression of regulated. Proc. Natl Acad. Sci. USA 94, 5955–5960
networks. Biophys. J. 75, 2240–2250 (1998). expansin modulates leaf growth and pedicel abscission in (1997).
41. Takeda, T. et al. Suppression and acceleration of cell Arabidopsis thaliana. Proc. Natl Acad. Sci. USA 97, 83. Brummell, D. A. et al. Modification of expansin protein
elongation by integration of xyloglucans in pea stem 9783–9788 (2000). abundance in tomato fruit alters softening and cell wall
segments. Proc. Natl Acad. Sci. USA 99, 9055–9060 Transgenic experiments to increase or reduce polymer metabolism during ripening. Plant Cell 11,
(2002). expansin gene expression, with results supporting 2203–2216 (1999).
42. Nishitani, K. & Tominaga, T. Endo-xyloglucan transferase, expansin’s role in cell growth and in abscission. Transgenic tomato fruits with higher levels of
a novel class of glycosyltransferase that catalyzes transfer 64. Choi, D., Lee, Y., Cho, H. T. & Kende, H. Regulation of LeEXPA1 gene expression were much softer than
of a segment of xyloglucan molecule to another xyloglucan expansin gene expression affects growth and controls, whereas fruits with reduced LeEXPA1
molecule. J. Biol. Chem. 267, 21058–21064 (1992). development in transgenic rice plants. Plant Cell 15, expression were firmer.
43. Steele, N. M. et al. Ten isoenzymes of xyloglucan 1386–1398 (2003). 84. Belfield, E. J., Ruperti, B., Roberts, J. A. & McQueen-
endotransglycosylase from plant cell walls select and 65. Zenoni, S. et al. Downregulation of the Petunia hybrida Mason, S. J. Changes in expansin activity and gene
cleave the donor substrate stochastically. Biochem. J. α-expansin gene PhEXP1 reduces the amount of expression during ethylene-promoted leaflet abscission in
355, 671–679 (2001). crystalline cellulose in cell walls and leads to phenotypic Sambucus nigra. J. Exp. Bot. 56, 817–823 (2005).
44. Purugganan, M. M., Braam, J. & Fry, S. C. The changes in petal limbs. Plant Cell 16, 295–308 (2004). 85. Cosgrove, D. J., Bedinger, P. & Durachko, D. M. Group I
Arabidopsis TCH4 xyloglucan endotransglycosylase. 66. Cho, H. T. & Kende, H. Expression of expansin genes is allergens of grass pollen as cell wall-loosening agents.
Substrate specificity, pH optimum, and cold tolerance. correlated with growth in deepwater rice. Plant Cell 9, Proc. Natl Acad. Sci. USA 94, 6559–6564 (1997).
Plant Physiol. 115, 181–190 (1997). 1661–1671 (1997). Shows that group-1 grass pollen allergens have
45. Yokoyama, R. & Nishitani, K. A comprehensive expression Shows that expression of certain EXPA genes in rice expansin activity and defines the second family of
analysis of all members of a gene family encoding cell-wall is increased upon submergence-induced internode expansins (β-expansins).
enzymes allowed us to predict cis-regulatory regions elongation. 86. Fry, S. C. et al. Xyloglucan endotransglycosylase, a new
involved in cell-wall construction in specific organs of 67. Reinhardt, D., Wittwer, F., Mandel, T. & Kuhlemeier, C. wall-loosening enzyme activity from plants. Biochem. J.
Arabidopsis. Plant Cell Physiol. 42, 1025–1033 (2001). Localized upregulation of a new expansin gene predicts 282, 821–828 (1992).
46. Rose, J. K., Braam, J., Fry, S. C. & Nishitani, K. The XTH the site of leaf formation in the tomato meristem. Plant Cell 87. Antosiewicz, D. M., Purugganan, M. M., Polisensky, D. H.
family of enzymes involved in xyloglucan 10, 1427–1437 (1998). & Braam, J. Cellular localization of Arabidopsis xyloglucan
endotransglucosylation and endohydrolysis: Current Shows that localized expression of an EXPA gene on endotransglycosylase- related proteins during
perspectives and a new unifying nomenclature. Plant Cell the flanks of the shoot apical meristem precedes development and after wind stimulation. Plant Physiol.
Physiol. 43, 1421–1435 (2002). emergence of the leaf primordium. 115, 1319–1328 (1997).
47. Thompson, J. E. & Fry, S. C. Restructuring of wall-bound 68. Brummell, D. A., Harpster, M. H. & Dunsmuir, P. Differential 88. Thompson, J. E. & Fry, S. C. Trimming and solubilization of
xyloglucan by transglycosylation in living plant cells. Plant J. expression of expansin gene family members during xyloglucan after deposition in the walls of cultured rose
26, 23–34 (2001). growth and ripening of tomato fruit. Plant Mol. Biol. 39, cells. J. Exp. Bot. 48, 297–305 (1997).
48. Strohmeier, M. et al. Molecular modeling of family GH16 161–169 (1999). 89. Redgwell, R. J. & Fry, S. C. Xyloglucan
glycoside hydrolases: potential roles for xyloglucan 69. Vriezen, W. H., De Graaf, B., Mariani, C. & endotransglycosylase activity increases during kiwifruit
transglucosylases/hydrolases in cell wall modification in Voesenek, L. A. C. J. Submergence induces expansin (Actinidia deliciosa) ripening (implications for fruit
the Poaceae. Protein Sci. 13, 3200–3213 (2004). gene expressin in flooding tolerant Rumex palustris and softening). Plant Physiol. 103, 1399–1406 (1993).
49. Schroder, R., Wegrzyn, T. F., Bolitho, K. M. & Redgwell, R. J. not in flooding intolerant R. acetosa. Planta 210, 956–963 90. Farkas, V., Sulova, Z., Stratilova, E., Hanna, R. &
Mannan transglycosylase: a novel enzyme activity in cell (2000). Maclachlan, G. Cleavage of xyloglucan by nasturtium seed
walls of higher plants. Planta 219, 590–600 (2004). 70. Lee, Y. & Kende, H. Expression of β-expansins is xyloglucanase and transglycosylation to xyloglucan subnit
50. Marga, F., Grandbois, M., Cosgrove, D. J. & Baskin, T. I. correlated with internodal elongation in deepwater rice. oligosaccharides. Arch. Biochem. Biophys. 298, 365–370
Cell wall extension results in the coordinate separation of Plant Physiol 127, 645–654 (2001). (1992).
parallel microfibrils: evidence from scanning electron 71. Wu, Y., Thorne, E. T., Sharp, R. E. & Cosgrove, D. J. 91. Fanutti, C., Gidley, M. J. & Reid, J. S. G. Action of a pure
microscopy and atomic force microscopy. Plant J. 43, Modification of expansin transcript levels in the maize xyloglucan endo-transglycosylase (formerly called
181–190 (2005). primary root at low water potentials. Plant Physiol. 126, xyloglucan-specific endo-1,4-β-D-glucanase) from the
51. Cosgrove, D. J. Rapid suppression of growth by blue light: 1471–1479 (2001). cotyledons of germinated nasturtium seeds. Plant J. 3,
occurrence, time course, and general characteristics. 72. Cho, H. T. & Cosgrove, D. J. Regulation of root hair 691–700 (1993).
Plant Physiol. 67, 584–590 (1981). initiation and expansin gene expression in Arabidopsis. 92. Bourquin, V. et al. Xyloglucan Endotransglycosylases have
52. Forterre, Y., Skotheim, J. M., Dumais, J. & Mahadevan, L. Plant Cell 14, 3237–3253 (2002). a function during the formation of secondary cell walls of
How the Venus flytrap snaps. Nature 433, 421–425 73. Chanliaud, E., de Silva, J., Strongitharm, B., Jeronimidis, vascular tissues. Plant Cell 14, 3073–3088 (2002).
(2005). G. & Gidley, M. J. Mechanical effects of plant cell wall 93. Matsui, A. et al. AtXTH27 plays an essential role in cell wall
53. Cosgrove, D. J. Wall extensibility: its nature, measurement, enzymes on cellulose/xyloglucan composites. Plant J. 38, modification during the development of tracheary
and relationship to plant cell growth. New Phytol. 124, 27–37 (2004). elements. Plant J. 42, 525–534 (2005).
1–23 (1993). 74. Cosgrove, D. J. Relaxation in a high-stress environment: 94. McQueen-Mason, S. J., Fry, S. C., Durachko, D. M. &
54. Cosgrove, D. J. Enzymes and other agents that enhance The molecular bases of extensible cell walls and cell Cosgrove, D. J. The relationship between xyloglucan
cell wall extensibility. Annu. Rev. Plant Physiol. Plant Mol. enlargement. Plant Cell 9, 1031–1041 (1997). endotransglycosylase and in-vitro cell wall extension in
Biol. 50, 391–417 (1999). 75. Shcherban, T. Y. et al. Molecular cloning and sequence cucumber hypocotyls. Planta 190, 327–331 (1993).
55. Rayle, D. L. & Cleland, R. E. Enhancement of wall analysis of expansins — a highly conserved, multigene 95. Cutillas-Iturralde, A. & Lorences, E. P. Effect of xyloglucan
loosening and elongation by acid solutions. Plant Physiol. family of proteins that mediate cell wall extension in plants. oligosaccharides on growth, viscoelastic properties, and
46, 250–253 (1970). Proc. Natl Acad. Sci. USA 92, 9245–9249 (1995). long-term extension of pea shoots. Plant Physiol. 113,
56. Cosgrove, D. J. Characterization of long-term extension of Cloning of α-expansin (EXPA) showed that it 103–109 (1997).
isolated cell walls from growing cucumber hypocotyls. belongs to a multigene family and lacks sequence 96. Kaku, T., Tabuchi, A., Wakabayashi, K., Kamisaka, S. &
Planta 177, 121–130 (1989). similarity to any enzymes known at the time of Hoson, T. Action of xyloglucan hydrolase within the native
Analyzes ‘acid-growth’ of isolated cell walls and publication. cell wall architecture and its effect on cell wall extensibility
hypothesizes a wall-loosening enzyme with unusual 76. McQueen-Mason, S. J. & Cosgrove, D. J. Expansin mode in azuki bean epicotyls. Plant Cell Physiol. 43, 21–26
properties, later purified and named expansin. of action on cell walls. Analysis of wall hydrolysis, stress (2002).
57. Hager, A., Menzel, H. & Krauss, A. Versuche und relaxation, and binding. Plant Physiol. 107, 87–100 (1995). 97. Fry, S. C. In vivo formation of xyloglucan nonasaccharide
hypothese zur primaerwirkung des auxins beim Detailed biochemical and biophysical analysis of — a possible biologically-active cell-wall fragment. Planta
streckungswachstum. Planta 100, 47–75 (1971) how α-expansins loosen plant cell walls. 169, 443–453 (1986).
(in German). 77. McQueen-Mason, S. & Cosgrove, D. J. Disruption of 98. Ito, H. & Nishitani, K. Visualization of EXGT-mediated
58. McQueen-Mason, S., Durachko, D. M. & Cosgrove, D. J. hydrogen bonding between plant cell wall polymers by molecular grafting activity by means of a fluorescent-
Two endogenous proteins that induce cell wall expansion proteins that induce wall extension. Proc. Natl Acad. Sci. labeled xyloglucan oligomer. Plant Cell Physiol. 40,
in plants. Plant Cell 4, 1425–1433 (1992). USA 91, 6574–6578 (1994). 1172–1176 (1999).

860 | NOVEMBER 2005 | VOLUME 6 www.nature.com/reviews/molcellbio

 REVIEWS

99. Fry, S. C. Novel ‘dot-blot’ assays for glycosyltransferases 112. Fry, S. C., Miller, J. G. & Dumville, J. C. A proposed role for 127. Cumming, C. M. et al. Biosynthesis and cell-wall
and glycosylhydrolases: optimization for xyloglucan copper ions in cell wall loosening. Plant Soil 247, 57–67 deposition of a pectin-xyloglucan complex in pea. Planta
endotransglycosylase (XET) activity. Plant J. 11, (2002). (2005) (doi:10.1007/s00425-005-1560-2).
1141–1150 (1997). 113. Fry, S. C. Oxidative scission of plant cell wall 128. Ye, Z. H. & Varner, J. E. Differential expression of two
100. Vissenberg, K., Fry, S. C. & Verbelen, J. P. Root hair polysaccharides by ascorbate-induced hydroxyl radicals. O-methyltransferases in lignin biosynthesis in Zinnia
initiation is coupled to a highly localized increase of Biochem. J. 332, 507–515 (1998). elegans. Plant Physiol. 108, 459–467 (1995).
xyloglucan endotransglycosylase action in Arabidopsis 114. Liszkay, A. & Schopfer, P. Plasma membrane-generated 129. Troughton, J. & Donaldson, L. A. Probing Plant Structure.
roots. Plant Physiol. 127, 1125–1135 (2001). superoxide anion radicals and peroxidase-generated (McGraw-Hill, New York 1972).
Uses fluorescent xylo-oligosaccharide to identify local hydroxyl radicals may be involved in the growth of 130. McCann, M. C., Wells, B. & Roberts, K. Direct visualization
sites of XET action, which in this case is site of the coleoptiles. Free Radic. Res. 37, 26–27 (2003). of cross-links in the primary plant cell wall. J. Cell Sci 96,
future bulge formation where a root hair is initiated. 115. Liszkay, A., Kenk, B. & Schopfer, P. Evidence for the 323–334 (1990).
101. Vissenberg, K. et al. Differential expression of AtXTH17, involvement of cell wall peroxidase in the generation of 131. Rizk, S. E., Abdel-Massih, R. M., Baydoun, E. A. &
AtXTH18, AtXTH19 and AtXTH20 genes in Arabidopsis hydroxyl radicals mediating extension growth. Planta 217, Brett, C. T. Protein- and pH-dependent binding of nascent
roots. Physiological roles in specification in cell wall 658–667 (2003). pectin and glucuronoarabinoxylan to xyloglucan in pea.
construction. Plant Cell Physiol. 46, 192–200 (2005). 116. Schopfer, P., Liszkay, A., Bechtold, M., Frahry, G. & Planta 211, 423–429 (2000).
102. Vissenberg, K., Fry, S. C., Pauly, M., Hofte, H. & Verbelen, Wagner, A. Evidence that hydroxyl radicals mediate auxin-
J. P. XTH acts at the microfibril-matrix interface during cell induced extension growth. Planta 214, 821–828 (2002). Acknowlegements
elongation. J. Exp. Bot. 56, 673–683 (2005). 117. Miller, J. G. & Fry, S. C. Characteristics of xyloglucan after I gratefully acknowledge the research contributions of numerous
103. Potter, I. & Fry, S. C. Changes in xyloglucan attack by hydroxyl radicals. Carbohydr. Res. 332, 389–403 students and colleagues, particularly D. M. Durachko, M. Perich,
endotransglycosylase (XET) activity during hormone- (2001). L.-C. Li, S. McQueen-Mason, T. Shcherban, S. Yuang, and Y. Wu.
induced growth in lettuce and cucumber hypocotyls and 118. Dumville, J. C. & Fry, S. C. Solubilisation of tomato fruit Our research is supported by grants from the National Science
spinach cell suspension cultures. J. Exp. Bot. 45, pectins by ascorbate: a possible non-enzymic mechanism Foundation, the National Institutes of Health and the Department
1703–1710 (1994). of fruit softening. Planta 217, 951–961 (2003). of Energy.
104. Hyodo, H. et al. Active gene expression of a xyloglucan 119. Ridley, B. L., O’Neill, M. A. & Mohnen, D. Pectins:
endotransglucosylase/hydrolase gene, XTH9, in structure, biosynthesis, and oligogalacturonide-related Competing interests statement
inflorescence apices is related to cell elongation in signaling. Phytochem. 57, 929–967 (2001). The author declares no competing financial interests.
Arabidopsis thaliana. Plant Mol. Biol. 52, 473–482 (2003). 120. Jones, L., Milne, J. L., Ashford, D. & McQueen-Mason, S. J.
105. Yokoyama, R., Rose, J. K. & Nishitani, K. A surprising Cell wall arabinan is essential for guard cell function.
diversity and abundance of xyloglucan Proc. Natl Acad. Sci. USA 100, 11783–11788 (2003). Online links
endotransglucosylase/hydrolases in rice. Classification 121. Zykwinska, A. W., Ralet, M. C., Garnier, C. D. &
and expression analysis. Plant Physiol. 134, 1088–1099 Thibault, J. F. Evidence for in vitro binding of pectin side DATABASES
(2004). chains to cellulose. Plant Physiol. 139, 397–407 (2005). The following terms in this article are linked online to:
106. Herbers, K., Lorences, E. P., Barrachina, C. & Sonnewald, U. Neutral pectins, such as arabinans, bind to cellulose TAIR: http://www.arabidopsis.org
Functional characterisation of Nicotiana tabacum surfaces, implying that pectin might crosslink CESA1 | CESA2 | CESA3 | CESA | CESA5 | CESA6 | KORRIGAN
xyloglucan endotransglycosylase (NtXET-1): generation of cellulose microfibrils.
transgenic tobacco plants and changes in cell wall 122. Baba, K., Sone, Y., Misaki, A. & Hayashi, T. Localization of FURTHER INFORMATION
xyloglucan. Planta 212, 279–287 (2001). xyloglucan in the macromolecular complex composed of CCRC Cell Wall Page:
107. Rose, J. K. C., Catala, C., Gonzalez-Carranza, Z. H. & xyloglucan and cellulose in pea stems. Plant Cell Physiol. http://www.ccrc.uga.edu/~mao/cellwall/main.htm
Roberts, J. A. The Plant Cell Wall (ed. Rose, J. K. C.) 35, 439–444 (1994). Cell Wall Genomics (Purdue):
264–324 (Blackwell, Oxford, 2003). 123. Hayashi, T., Ogawa, K. & Mitsuishi, Y. Characterization of http://cellwall.genomics.purdue.edu
108. Libertini, E., Li, Y. & McQueen-Mason, S. J. Phylogenetic the adsorption of xyloglucan to cellulose. Plant Cell WallBioNet (MSU):
analysis of the plant endo-β-1,4-glucanase gene family. Physiol. 35, 1199–1205 (1994). http://xyloglucan.prl.msu.edu
J. Mol. Evol. 58, 506–515 (2004). 124. Park, Y. W. et al. Enhancement of growth by expression of Carbohydrate-Active EnZYmes:
109. Ohmiya, Y. et al. Evidence that endo-1,4-β-glucanases act poplar cellulase in Arabidopsis thaliana. Plant J. 33, http://afmb.cnrs-mrs.fr/CAZY
on cellulose in suspension- cultured poplar cells. Plant J. 1099–1106 (2003). Cellulose synthase-like:
24, 147–158 (2000). 125. Keegstra, K., Talmadge, K. W., Bauer, W. D. & Albersheim, P. http://cellwall.stanford.edu/cesa/index.shtml
110. Ohmiya, Y., Takeda, T., Nakamura, S., Sakai, F. & Hayashi, T. The structure of plant cell walls. III. A model of the walls of XTH: http://labs.plantbio.cornell.edu/XTH/arabidopsis.htm
Purification and properties of a wall-bound endo-1, 4-β- suspension-cultured sycamore cells based on the Expansins: http://www.bio.psu.edu/expansins
glucanase from suspension-cultured poplar cells. Plant interconnections of the macromolecular components. Daniel Cosgrove’s laboratory:
Cell Physiol. 36, 607–614 (1995). Plant Physiol. 51, 188–196 (1973). http://www.bio.psu.edu/people/faculty/cosgrove
111. Tsabary, G. et al. Abnormal ‘wrinkled’ cell walls and 126. Thompson, J. E. & Fry, S. C. Evidence for covalent
retarded development of transgenic Arabidopsis thaliana linkage between xyloglucan and acidic pectins in SUPPLEMENTARY INFORMATION
plants expressing endo-1,4-β-glucanase (cell) antisense. suspension-cultured rose cells. Planta 211, 275–286 See online article: S1 (figure)
Plant Mol. Biol. 51, 213–224 (2003). (2000). Access to this interactive links box is available online.

NATURE REVIEWS | MOLECULAR CELL BIOLOGY VOLUME 6 | NOVEMBER 2005 | 861