Directory UMM :Data Elmu:jurnal:P:PlantScience:Plant Science_BioMedNet:061-080:
trends in plant science
Perspectives
The role of root border cells
in plant defense
Martha C. Hawes, Uvini Gunawardena, Susan Miyasaka
and Xiaowen Zhao
The survival of a plant depends upon the capacity of root tips to sense and move
towards water and other nutrients in the soil. Perhaps because of the root tip’s
vital role in plant health, it is ensheathed by large populations of detached
somatic cells – root ‘border’ cells – which have the ability to engineer the
chemical and physical properties of the external environment. Of particular
significance, is the production by border cells of specific chemicals that can
dramatically alter the behavior of populations of soilborne microflora. Molecular
approaches are being used to identify and manipulate the expression of plant
genes that control the production and the specialized properties of border cells
in transgenic plants. Such plants can be used to test the hypothesis that these
unusual cells act as a phalanx of biological ‘goalies’, which neutralize dangers
to newly generated root tissue as the root tip makes its way through soil.
E
ach day, thousands of uniquely differentiated root border cells are synthesized
and programmed to separate from the
root tips (the region at the apex housing the apical meristem and root cap) of higher plants1,2.
Although they are attached to the root periphery
by a water soluble polysaccharide matrix, the
middle lamellae of these cells become solubilized by the action of pectolytic enzymes in the
cell wall3. By definition, border cells are those
cells that disperse into suspension within seconds when root tips are placed into water (Fig.
1). These cells originate from root cap meristematic cells, which give rise to cell layers that
progressively differentiate through specialized
stages before they finally separate from the cap
periphery4. Thus, by the time an individual border cell is delivered to its ultimate destination
external to the root, it has already served several
functions, such as mucilage secretion, the sensing of gravity and other environmental signals.
How many of those functions continue to operate in border cells is unknown. However, the
border cells of most species are viable after they
detach from the root (Fig. 1), and in culture the
cells can divide and differentiate into organized
tissue (Fig. 1). Upon separation from the cap,
metabolic activity in border cells increases and
gene expression undergoes a global switch, such
that mRNA and protein profiles are grossly distinct from those of progenitor cells in the root
cap5. Border cells then produce specific metabolites, such as anthocyanins and antibiotics6
(Fig. 1), and enzymes including a Rhizobiuminduced peroxidase7 and a low-pH galactosidase
(C. Jian and M.C. Hawes, unpublished; Fig. 1).
We call these unusual cells (previously termed
‘sloughed root cap cells’) root ‘border’ cells to
128
March 2000, Vol. 5, No. 3
emphasize that they constitute a biotic boundary layer between the root surface and the soil.
In the past, border cells largely have
been overlooked as a structural and functional
component of root systems. One reason for this
oversight is the notion that the cells are a moribund by-product of root-cap turnover. This presumption has prevailed in recent years in spite
“…environmental
signals
can override the control of
border cell development using
endogenous signals.”
of the fact that L. Knudson8 demonstrated that
it was incorrect in 1919 (Refs 1,2). Although
species in a few families, such as the
Compositae, have border cells that die before
detachment, in the majority of species examined, border cell populations are .90% viable5.
Border cells of maize and pea grown hydroponically can remain viable for more than three
months8, and cells of maize have been reported
to survive for a week or more in field soil9. A
second reason for the lack of attention paid to
border cells is that the cells disperse so efficiently in response to free water that even the
gentlest washing of root systems to remove soil
before experimental manipulation and examination removes all border cells1. However, in the
absence of free water the cells cling so tenaciously to the root periphery that their presence
is difficult to distinguish even by experienced
investigators.
Regulation of border cell production
The number of border cells that can be produced daily by a given root is conserved at the
family level, and can range from a dozen for
tobacco to .10 000 for cotton and pine. The
process of border cell production for tap roots,
lateral and branch roots appears to be similar.
Therefore, for complex root systems with hundreds of branches, millions of border cells
would be deposited into the soil daily if their
production were a constitutively expressed
process, which was assumed to be the case for
many years10. However, recent studies have
established that border cell production in
species such as cereals, legumes and cotton, is
a tightly regulated process that is controlled by
endogenous and environmental signals. In the
laboratory, radicles synthesize a set number of
border cells within 24 h of emergence. As they
accumulate on the cap periphery, border cells
of pea release an extracellular suppressor that
specifically inhibits mitosis in the root-cap
meristem, without affecting mitosis in the
adjacent apical meristem that gives rise to root
growth11. Root-cap turnover ceases and no
new border cells are made, but the set of
~4000 cells that has been made remains
appressed to the periphery of the cap as the
root elongates. If border cells are removed, or
if the extracellular suppressor is diluted by
dipping the root into water, cell division in the
root cap meristem resumes within 5 min, and
remains high for 2 h. New cells can be collected from the cap periphery within 1 h after
removal of the old ones; after 24 h a new set
of 4000 cells is complete and no more cells are
made. This process, which can be synchronized experimentally, has been exploited to
identify two genes, psugt1 and rcpme1, which
play a role at both ends of border cell development – cell division in the cap meristem and
cell separation at the cap periphery, respectively3,12. When the expression of either gene
in transgenic roots is inhibited by antisense
mRNA mutagenesis, border cell development
is blocked. Such mutants are being used under
controlled conditions to dissect the molecular
and cellular mechanisms by which border cell
development is regulated.
How border cell development is regulated
in soil is unknown. Because border cells do not
detach readily except when they are actually
placed into water, a root growing through conditions where the tip never experiences a flush
of water, could, theoretically, be expected to
retain the same set of border cells throughout
its lifetime. Renewed border cell production in
response to any exposure to free water, such as
during rain or irrigation, would be a logical
prediction. However, when roots are grown in
a wet semisolid matrix, such as water agar,
where border cells can be seen to be removed continuously when viewed microscopically, the results are not always found to be
1360 - 1385/00/$ – see front matter © 2000 Elsevier Science Ltd. All rights reserved. PII: S1360-1385(00)01556-9
trends in plant science
Perspectives
consistent with such straightforward predictions. Under these conditions, the extracellular suppressor, which inhibits cell division in
the root cap meristem, presumably would be
diluted continuously or would be removed by
physical forces as the tip elongates, thus a constant induction of mitosis with a continuous
production of border cells would be expected.
In some cases, continuous production does
occur, such that the entire root is ensheathed
in layers of detached border cells. In other
instances, sometimes in adjacent roots in a single Petri dish, border cell production occurs
only intermittently or not at all, suggesting that
localized conditions around individual roots
might influence the process1. Recent studies
have confirmed that environmental signals
can override the control of border cell development using endogenous signals13. For
example, when pea is exposed to increased
carbon dioxide levels, the normal control of
border cell production is overridden, such that
instead of producing the usual set of 4000
cells, twice as many cells are made (Fig. 2).
By contrast, border cell development in alfalfa
is impervious to increased carbon dioxide.
The regulation of this process at multiple
levels, with species-dependent variation in
responses to external signals, is one of the
properties that led us to suggest that border
cells can act as cellular ‘goalies’ (players
assigned to protect the goal in various sports;
Webster’s Unabridged Dictionary) to ward
off threats to the apical meristem.
Proposed functions of root border cells
The root cap provides physical protection for
the apical meristem; the border cells and their
associated mucilage were long presumed to
provide lubrication for the passage of the root
cap. To date, the hypothesis that border cells
lubricate the root tip is unsupported by experimental data2,14,15. The fact that plants turn
border cell production on and off without
hampering their ability to penetrate solid
media, and that a few species do not appear to
undergo border cell production at all argues
against such a ‘housekeeping’ function1,2 . We
have proposed that border cells serve plants by
enabling roots to define their own ecology.
The best characterized properties of border
cells involve their specific recognition and
responsiveness to soilborne microflora1,2. In a
genotype-dependent manner, border cells can:
• Attract zoospores (Fig. 1).
• Synthesize defensive structures in response
to fungal attack16.
• Repel or bind pathogenic bacteria (Fig. 1).
• Control growth and gene expression in
symbiotic bacteria (Fig. 1).
Chemotactic attraction to border cells of soilgrown plants has been shown to facilitate root
infection by pathogenic bacteria, and such
recognition might create ‘biased rhizospheres’
(b)
(a)
(h)
(c)
(g)
(d)
(e)
(f)
Fig. 1. Properties of root border cells. (a) Border cells have dispersed away from the cotton root
tip after immersion in water for 30 sec, showing one days’ accumulation of border cells
(~10 000) surrounding the tip (scale bar 5 1 mm). (b) The viability of detached border cells is
revealed by the accumulation of a vital stain, fluorescein diacetate (scale bar 5 10 mm). (c)
Cell division in border cells – induced by incubation in tissue culture medium containing plant
hormones28 (scale bar 5 10 mm). (d) Expression of the red-pigmented antibiotic shikonin is
localized in root border cells of Lithospermum erythrorhizon6 (scale bar 5 1 mm). (e)
Expression of a low-pH galactosidase is specific to root border cells of pea (C. Jian and M.C.
Hawes, unpublished) (scale bar 5 1 mm). (f) Border cells of cotton specifically attract
zoospores of Pythium dissotocum, which germinate, penetrate and kill the cells within 2 min
(Ref. 29) (scale bar 5 10 mm). (g) Genotype-specific binding of Agrobacterium tumefaciens
by border cells is correlated with susceptibility to infection19 (scale bar 5 10 mm). (h) When
pea roots are incubated on a lawn of Rhizobium leguminosarum expressing a nod-lacZ reporter
gene on solidified culture medium (pH 7.0), intense nod gene expression is detected in bacteria present at the root tip in the region where border cells are present30,31 (scale bar 5 1 mm).
by dictating which populations can colonize
plant root systems17. By modulating the properties of the external environment surround-
“…border cells can act as
cellular ‘goalies’…to ward off
threats to the apical meristem.”
ing the advancing root tip, the regulated
production of border cells could, directly or
indirectly, influence virtually all of the physical and chemical parameters (i.e. pH, nutrient
solubility or soil structure) of the root–
soil ecosystem. Here we focus on the direct
role that border cells might play in defense.
The recently discovered cellular responses of
border cells to nematodes, aluminum and
fungi suggest novel mechanisms by which
border cells might function to protect the root
tip from biotic and abiotic stress.
Border cells attract and immobilize
nematodes
In Greek mythology, the Sirens’ beautiful
songs tempted sailors to linger on their island,
and thus distracted them forever from completing their intended journeys. Recent studies
suggest that border cells might carry out a
similar distraction to pathogenic nematodes on
their way to infect root tips (X. Zhao and M.C.
Hawes, unpublished). When a pea root, whose
border cells have been removed, is placed onto
water agar containing a lawn of root knot nematodes, no sign of recognition is evident at the
root tip (Fig. 3). By contrast, when border cells
are present nematodes rapidly accumulate
around the root tip periphery (Fig. 3). When
March 2000, Vol. 5, No. 3
129
trends in plant science
Perspectives
are consistent with the hypothesis that border
cells have the capacity to synthesize an inducible extracellular structure that interferes
with the ability of aluminum to cause further
cellular damage. The mechanism by which this
occurs is unknown, but one possibility is that
the charged aluminum molecule becomes
immobilized by binding to the polysaccharide
layer20. If this is the case, the capacity of thousands of detached border cells to immobilize
aluminum ions could have a profound effect on
the ability of aluminium to damage the root tip,
the primary site of aluminum’s toxic effects21.
Border cells act as a decoy for
fungal infection
Fig. 2. Stimulation of border cell production by carbon dioxide. (a) In an ambient atmosphere,
roots of germinating seedlings produce ~4000 border cells in 24 h, cell production then ceases
unless the existing cells are removed. (b) In atmospheres containing increased carbon dioxide,
border cell production does not level off, and more than twice as many cells accumulate13.
these roots are lifted from the surface, clumps
of border cells left behind reveal high populations of actively motile nematodes (Fig. 3).
However, within 30 min, the motility of the
nematodes ceases and the worms appear
straight and inert, as they do when dead (Fig. 3).
When the immobilization phenomenon is measured using quantitative assays, a .99% loss of
mobility develops in response to a single heat-
stable, polar fraction of the root exudates (soluble extracellular material). This immobilization effect is reversible – within a few hours to
a few days (depending on conditions of the
assay) the nematodes resume full mobility. If
a similar process occurs in the soil, by the time
that the nematodes resume motility, a root
tip growing at a rate of 1 mm per hour
would be long past being in danger of penetration. This phenomenon might account, in
part, for the fact that the normal site of infection
by nematodes is behind the root tip, just past the
region where border cells are released18.
When pea seedlings are inoculated uniformly
with spores of pathogenic fungi, nearly all
develop visible lesions in the region just behind
the root tip within 24 h (U. Gunawardena and
M.C. Hawes, unpublished). A small proportion of seedlings also develop a visible lesion
at the root tip. When excised and placed onto
culture medium, fungal mycelium emerges
from these infected tips, confirming that the
root tip is not inherently resistant to infection
(Fig. 5). In the majority of seedlings (.90%),
the root tip remains white, and to the naked
eye appears to be infection free. However,
microscopic observation reveals that the tip is
actually covered in fungal hyphae (Fig. 5).
Mucilage production can repel bacteria
and decrease sensitivity to aluminum
Fig. 3. Attraction and immobilization of
the root knot nematode by root border
cells. (a) When a pea root is placed onto
water agar inoculated with nematodes,
nematodes are rapidly attracted to surround the border cells (right), but do not
surround roots whose border cells have
been removed (left). (b). When roots are
lifted from the surface, the active motility
of nematodes can be detected within
clumps of border cells left behind. (c)
After 30 min, motility within the clumps
of border cells ceases, and the nematodes
are rigid and inert.
130
March 2000, Vol. 5, No. 3
Border cells of legumes and cereals produce a
mucilage layer in response to co-cultivation
with pathogenic bacteria, but not in response
to E. coli19. This layer appears to repel the bacteria (Fig. 4), but its mechanism of action is
unknown. Recent studies have revealed that
co-cultivation of border cells with aluminum
results in the production of a mucilage layer
similar to that which occurs in response to
pathogenic bacteria (S. Miyasaka and M.C.
Hawes, unpublished). Within 2 h of exposure
to aluminum, a layer of mucilage around individual border cells increases in a dosagedependent manner, from being barely
detectable (Fig. 4), to being as wide (Fig. 4) or
wider than the cell’s diameter (Fig. 4).
Interestingly, the development of the layer is
correlated with a near-total cessation of aluminum-induced border cell death. For the first
2 h of exposure to aluminum, a dosage-dependent linear-rate of border cell death proceeds.
After 2 h, when the mucilage layer is in place,
the rate of cell death drops precipitously, such
that the viability of the cell population remains
largely static for the next 20 h (S. Miyasaka
and M.C. Hawes, unpublished). The results
Fig. 4. Development of a mucilage
layer by individual border cells. (a) A border cell of oats incubated for 2 h with
Agrobacterium tumefaciens has developed a mucilage layer that excludes the
bacteria, which can be seen at high density
outside the layer. (b) A border cell incubated in water and placed into India ink (a
stain that is excluded by polysaccharide)
exhibits a narrow layer of mucilage
around its surface. Border cells incubated
for 2 h in (c) 100 mM aluminum or in (d)
200 mM aluminum develop layers as wide
as or wider than the diameter of the cell.
trends in plant science
Perspectives
tear-away jerseys that were designed to detach
when grabbed by would-be tacklers, leaving
the runner free to keep moving down the field.
A similar mechanism, mediated by border
cell detachment, could, in part, account for
the observation that root tips generally escape
infection and colonization in the field2.
Model
Fig. 5. Mantle production in response to
pathogenic fungi (U. Gunawardena et al.,
unpublished). When roots are inoculated
with spores of Nectria haematococca,
most root tips remain white and appear
to be uninfected, but a few seedlings
develop visible lesions at the tip. When
such infected tips are plated onto culture
medium, abundant fungal growth
emerges (a, left). Most root tips remain
white and appear to be uninfected, but
when examined microscopically, are seen
to be covered in a sheath or ‘mantle’ of
fungal hyphae (b). When these roots are
placed into water and agitated gently, the
mantle of fungus falls away from the
root tip. When the tip is excised and
placed onto culture medium, no fungal
growth emerges (a, center), indicating
that the tips are sterile like the tips of the
uninoculated control seedlings (a, right).
When the tip is placed into water and agitated
gently, this ‘mantle’ detaches and can be seen
microscopically to consist of thousands of
digested border cells held together by actively
growing fungal mycelium2. Even under extreme
conditions, which include high concentrations
of spores and heat-shock temperatures, the root
tip remains sterile for days (Fig. 5), indistinguishable from uninoculated controls (Fig. 5) in
spite of having been covered by a sheath of
necrogenic fungus. In these plants, root growth
and development are comparable to that of the
controls, indicating that the apical meristems are
fully functional. In the 1950s, running backs
(American football players) were outfitted with
Taken together, the observed responses of border cells to aluminum, nematodes and pathogenic fungi are consistent with a twostep model in which the root tip is protected
by an unusual and complex process. First, the
root cap meristem actively produces a ‘front’
consisting of thousands of border cells that can
engineer the environment to render a threat
temporarily harmless. Elongating cells generated by the apical meristem then push the root
tip into and through the newly ‘engineered’
environment. The particular mechanism by
which the threat is rendered harmless by the
border cells might vary and could be constitutive (i.e. the secretion of a chemical that anesthetizes nematodes) or inducible in response
to a particular signal (i.e. the production of a
mucilage layer that immobilizes aluminum).
It is important to note again that only in the
presence of a flush of free water, such as after
rain or irrigation, would the en masse release
“…the capacity of thousands
of detached border cells to
immobilize aluminum ions
could have a profound effect
on the ability of aluminium
to damage the root tip, the
primary site of aluminum’s
toxic effects.”
of border cell populations occur. In the
absence of water, existing border cells remain
appressed to the root periphery and are carried
along as the root grows. It is also only in the
presence of free water that microorganisms or
toxins, such as aluminum, constitute a threat
to plants, because only then can movement
and uptake occur.
Reasons for protecting the tip in a
different way
The root tip is essential to plant health
This costly defensive process might be needed
because of the multifaceted role played by tissues within the root tip. One of the most-cited
functional distinctions between animals and
plants is that animals can move and plants cannot. Whereas animals can respond to signals
to avoid danger and seek nutrients, it is said
that plants must rely on internal signals to
Fig. 6. Directed movement of roots in
response to exposure to aluminum. Pea
seedlings with 1-cm-long roots were
placed onto medium containing no aluminum (left), 100 mM (center) or 200 mM
aluminum (right). Arrows indicate the
position of the root tip at the time that the
roots were placed onto the medium. In
response to aluminum, roots immediately began to grow away from the
plate at a 458 angle (S. Miyasaka and
M.C. Hawes, unpublished).
generate chemical responses in situ to synthesize their own food and protect themselves
from stresses. This simple dogma does not
take into account that complex root systems,
with each root tip elongating at a rate of 1 mm
per hour, can traverse .6 m of soil in a day22.
As in animals, this movement occurs as a
result of specific signals perceived by root
cells and transduced into behavioral responses
via common gene expression systems. Plant
roots move towards nutrients and water, and
away from toxins, such as aluminum (Fig. 6),
by sensing signals that notify them of the
presence of such stimuli. Anyone who has
ever battled to keep willow (Salix) roots out of
their septic tanks will appreciate the effectiveness of roots’ directed-movement capacities. Signals generated by the root as a result
of the detection of toxins or the uptake of
nutrients, are transmitted to the aboveground
parts of the plant. This communication
forms the basis, not only for the daily metabolism of the plant and its ultimate reproductive activities, but also for the overall size and
architecture of the whole plant23.
The primary site for perception of underground signals is the root tip23. Cells within the
terminal few millimeters of the root respond
to light, electricity, water, touch, ions and
other soluble molecules, to generate impulses
that direct the rate, direction and magnitude of
root movement. Just like animals, a capacity
to move away from danger and towards water
and other required elements is a crucial
component of plant life. So perhaps a more
important distinction between plants and animals is the mechanism by which movement is
generated: whereas animals can get up and
walk, plants can move only by the generation
and elongation of new cells by root meristems.
March 2000, Vol. 5, No. 3
131
trends in plant science
Perspectives
This newly synthesized tissue comprises the
same terminal few millimeters of tissue that
detects and responds to environmental stimuli
to condition movement toward nutrients and
away from danger. Not surprisingly, such
newly synthesized tissue is also especially
vulnerable to abiotic and biotic disease. For
humans, this is the functional equivalent of
having to rely on newborn babies to explore
and search a predator-filled wilderness for
supplies to send back to base camp. This biological dilemma might help to account for the
fact that in most species the vulnerable but
essential root meristems are surrounded by an
army of border cell ‘foot soldiers’2.
Other defensive strategies might not work
for the tip
We have proposed that a fail-safe defense
mechanism is needed to protect the root tip
because of its importance to the overall health
of the plant. However, it is well established
that plants already have complex defense
pathways that are activated rapidly in response
to external signals. Within moments of recognition of biotic and abiotic stresses, plants can
alter their surface structures, generate toxic
moieties, and activate the expression of genes
for the synthesis of dozens of antimicrobial
proteins and other metabolites24. Why would
an elaborate alternative method involving the
separation of masses of detached cells be
needed by plants that already are well
equipped to defend their cells? One explanation is that defense responses in plants often
involve the so-called ‘hypersensitive reaction’, in which the rapid necrosis of a few plant
cells is thought to inhibit the spread of a
pathogen25. In most tissues, including the
region behind the root tip where most
root–microbe relationships are initiated, the
sacrifice of a few cells offers containment of
the infection with little or no deleterious effect
on plant health. However, in the root tip, there
is no margin for the loss of a few cells in the
interest of containing infection when those
few cells comprise the apical meristem.
Indeed, experiments have confirmed that even
a minor necrotic reaction in the terminal 2 mm
of the root tip is correlated with the rapid cessation of growth of that root (U. Gunawardena
and M.C. Hawes, unpublished).
The activation of defense pathways in
response to biotic and abiotic signals does not
always result in hypersensitive cell death26.
The production of antimicrobial chemicals
can occur systemically, without any association with cell death. However, even in
the absence of cell death the activation of
defense pathways is a costly metabolic
process. As with other such responses (e.g.
heat shock), plants cope with this cost by
down-regulating other pathways. One activity that is known to be virtually shut down in
132
March 2000, Vol. 5, No. 3
response to the activation of defense pathways is the expression of genes needed for
cell division27. In most tissues, this would not
present a problem, but in a meristem it does.
In a meristem that plays a central role in
allowing the root system to avoid potentially
lethal stresses, a transient inhibition of
cell division during exposure to pathogens or
toxins could be deadly.
Future perspectives
Our model is congruent with the known biology of root infection: nearly all pathogenic
and symbiotic relationships are initiated not
“…in the root tip, there is no
margin for the loss of a few
cells in the interest of containing infection when those
few cells comprise the apical
meristem.”
at the tip, but in the region of elongation, just
behind the tip where the production of border
cells does not occur. We suspect that the main
reason that the effect of border cells on root
infection has never been examined, is its
effectiveness: so few pathogens and symbionts have been reported to initiate infection
or colonization at the root tip that there has
been little reason to focus on its role in infection. The availability of plants with altered
border cell production means that the most
obvious prediction of our model can now be
tested. If the presence of border cells is the
reason root tips are largely impervious to
infection, then inhibiting their production
and/or release will result in a change in the
site and/or nature of infection. If this is the
case, the study of how border cells can defend
the tip so successfully might elucidate novel
defense strategies that can be applied to
tissues that do not come equipped with
their own cellular ‘goalies’. It might not be
feasible to program other tissues to produce
and separate populations of border cells.
However, genetic engineering approaches
can be used to harness the cellular and extracellular mechanisms by which border cells
neutralize dangerous molecules or organisms
and express them in tissues that are not as well
protected as the root tip.
Acknowledgements
The authors thank Merritt R. Nelson for
the ‘tear-away jersey’ model for border cell
function, Hans D. VanEtten for critical reading of the manuscript, and the Dept of Energy,
Division of Energy Biosciences and United
States Dept of Agriculture for funds supporting basic research on the functions of root
border cells.
References
1 Hawes, M.C. and Brigham, L.B. (1992) Impact
of root border cells on microbial populations in
the rhizosphere. Adv. Plant Pathol. 8, 119–148
2 Hawes, M.C. et al. (1998) Function of root
border cells in plant health: pioneers in the
rhizosphere. Annu. Rev. Phytopathol. 36, 311–327
3 Wen, F. et al. (1999) Effect of pectin
methylesterase gene expression on pea root
development. Plant Cell 11, 1129–1140
4 Feldman, L.J. (1984) Development and dynamics
of the root apical meristem. Am. J. Bot. 7,
1308–1314
5 Brigham, L.A. et al. (1995) Differential
expression of proteins and mRNAs from border
cells and root tips of pea. Plant Physiol. 109,
457–463
6 Brigham, L.A. et al. (1999) Cell-specific
production and antimicrobial activity of
naphthoquinones in roots of Lithospermum
erythrothizon. Plant Physiol. 119, 417–428
7 Cook, D. et al. (1995) Transient induction of a
peroxidase gene in Medicago truncatula precedes
infection by Rhizobium meliloti. Plant Cell 7, 43–55
8 Knudson, L. (1919) Viability of detached root
cap cells. Am. J. Bot. 6, 309–310
9 Vermeer, J. and McCully, M.E. (1982) The
rhizosphere in Zea: new insight into its structure
and development. Planta 156, 45–61
10 Clowes, F.A.L. (1994) Origin of the epidermis in
root meristem. New Phytol. 127, 335–347
11 Brigham, L.A. et al. (1998) Meristem-specific
suppression of mitosis and a global switch in
gene expression in the root cap of pea by
endogenous signals. Plant Physiol. 118, 1223–1231
12 Woo, H.H. et al. (1999) Meristem-localized
inducible expression of a UDPglycosyltransferase gene is essential for plant
growth and development in pea and alfalfa. Plant
Cell 11, 2303–2315
13 Zhao, X. et al. (2000) Stimulation of border cell
production in response to increased CO2 levels.
Plant Physiol. 122, 181–188
14 Rogers, H.T. et al. (1942) The source and
phosphatase activity of exoenzyme systems of
corn and tomato roots. Soil Sci. 54, 353–365
15 Guinel, F.C. and McCully, M.E. (1986) Some
water-related physical properties of maize root
cap mucilage. Plant Cell Environ. 9, 657–666
16 Sherwood, R. (1987) Papilla formation in corn
root cap cells and leaves inoculated with
Colletotrichum graminicola. Phytopathology 77,
930–934
17 Hawes, M.C. and Smith, L. (1988) Requirement
for chemotaxis in pathogenicity of Agrobacterium
tumefaciens on roots of soil-grown pea plants.
J. Bacteriol. 171, 5668–5671
18 Prot, J.C. (1980) Migration of plant-parasitic
nematodes towards plant roots. Revue Nematol.
3, 305–318
19 Hawes, M.C. and Pueppke, S.G. (1987)
Correlation between binding of Agrobacterium
tumefaciens by root cap cells and susceptibility of
plants to crown gall. Plant Cell Rep. 6, 287–290
trends in plant science
Perspectives
20 Horst, W.J. et al. (1982) Mucilage protects root
meristem from aluminum injury.
Z. Pflanzenphysiol. Bd. 105, 435–444
21 Ryan, P.R. et al. (1995) Characterisation of
Al-stimulated efflux of malate from the apices of
Al-tolerant wheat roots. Planta 196,
103–110
22 Lynch, J. and van Beem, J.J. (1993) Growth and
architecture of seedling roots of common bean
genotypes. Crop Sci. 33, 1253–1257
23 Aiken, R.M. and Smucker, A.J.M. (1996) Root
system regulation of whole plant growth. Annu.
Rev. Phytopathol. 34, 325–346
24 Somssich, I.E. and Hahlbrock, K. (1998)
Pathogen defence in plants – a paradigm of
biological complexity. Trends Plant Sci. 3, 86–90
25 Kamoun, S. et al. (1999) Resistance to
oomycetes: a general role for the hypersensitive
response? Trends Plant Sci. 4, 196–200
26 Grant, M. and Mansfield, J. (1999) Early events
in host–pathogen interactions. Curr. Opin. Plant
Biol. 2, 312–319
27 Logemann, E. et al. (1995) Gene activation by
UV light, fungal elicitor or fungal infection in
Petroselinum crispum is correlated with
repression of cell cycle-related genes. Plant J. 8,
865–876
28 Hawes, M.C. and Pueppke, S.G. (1986) Sloughed
peripheral root cap cells: yield from different
species and callus formation from single cells.
Am. J. Bot. 73, 1466–1473
29 Goldberg, N. et al. (1989) Specific attraction to
and infection of cotton root cap cells by
zoospores of Pythium dissotocum. Can. J. Bot.
67, 1760 –1767
30 Zhu, Y. et al. (1997) Release of nodulation gene
inducing chemicals from root border cells. Plant
Physiol. 115, 1691–1698
31 Peters, N.K. and Long, S.R. (1988) Alfalfa
root exudates and compounds which promote
or inhibit induction of Rhizobium meliloti
nodulation genes. Plant Physiol. 88,
396–400
Martha C. Hawes*, Uvini Gunawardena
and Xiaowen Zhao are at the University
of Arizona, Dept of Plant Pathology,
Tucson, AZ 85721, USA;
Susan Miyasaka is at the University
of Hawaii, 461 W. Lanikaula St, Hilo,
HI 96720, USA.
*Author for correspondence (tel 11 520 621
5490; fax 11 20 621 9290;
e-mail mhawes@u.arizona.edu).
Errata
The review article by Richard A. Dixon and Christopher L. Steele (1999) Flavonoids and isoflavonoids – a gold mine for metabolic engineering. Trends Plant Sci. 4, 394–400 in the October 1999 issue of Trends in Plant Science contained two errors.
In Figure 2, dihydroflavonol (not flavanone) is acted on by flavonol synthase (FLS) to form flavonol. In Figure 4, the structure of the 2-hydroxyisoflavanone intermediate was depicted incorrectly. Corrected versions of these portions of the figures are
printed below.
OH
HO
OH
O
HO
OH
OH
(R1)
FLS
OH
OH
O
Flavanone
PII:S1360-1385(00)1572-7
H
(OH)
O
HO
O
HO
F3βH
O
Flavonol
O 2
O
OH
2-Hydroxyisoflavanone
OH
OH
O
Dihydroflavonol
Fig. 2.
Fig. 4.
Erratum
phyAr
phyAfr
Light
PII:S1360-1385(00)1573-9
phyBr
Fig. 1.
phyBfr
In the December 1999 issue of Trends in
Plant Science, we published a research
news article by Albrecht G. von Arnim
(1999) Phytochrome in the limelight.
Trends Plant Sci. 4, 465–466. In Figure 1,
top left it should read phyAr and not
phyBr. A corrected version of this portion
of the figure is printed left.
We apologize to the authors and to our
readers for these errors. A corrected version of these articles can be viewed in
HTML format at plants.trends.com
Students
Did you know that you are
entitled to a
50% discount on a
subscription to
Trends in Plant Science?
See the subscription order
card in this issue
for details.
March 2000, Vol. 5, No. 3
133
Perspectives
The role of root border cells
in plant defense
Martha C. Hawes, Uvini Gunawardena, Susan Miyasaka
and Xiaowen Zhao
The survival of a plant depends upon the capacity of root tips to sense and move
towards water and other nutrients in the soil. Perhaps because of the root tip’s
vital role in plant health, it is ensheathed by large populations of detached
somatic cells – root ‘border’ cells – which have the ability to engineer the
chemical and physical properties of the external environment. Of particular
significance, is the production by border cells of specific chemicals that can
dramatically alter the behavior of populations of soilborne microflora. Molecular
approaches are being used to identify and manipulate the expression of plant
genes that control the production and the specialized properties of border cells
in transgenic plants. Such plants can be used to test the hypothesis that these
unusual cells act as a phalanx of biological ‘goalies’, which neutralize dangers
to newly generated root tissue as the root tip makes its way through soil.
E
ach day, thousands of uniquely differentiated root border cells are synthesized
and programmed to separate from the
root tips (the region at the apex housing the apical meristem and root cap) of higher plants1,2.
Although they are attached to the root periphery
by a water soluble polysaccharide matrix, the
middle lamellae of these cells become solubilized by the action of pectolytic enzymes in the
cell wall3. By definition, border cells are those
cells that disperse into suspension within seconds when root tips are placed into water (Fig.
1). These cells originate from root cap meristematic cells, which give rise to cell layers that
progressively differentiate through specialized
stages before they finally separate from the cap
periphery4. Thus, by the time an individual border cell is delivered to its ultimate destination
external to the root, it has already served several
functions, such as mucilage secretion, the sensing of gravity and other environmental signals.
How many of those functions continue to operate in border cells is unknown. However, the
border cells of most species are viable after they
detach from the root (Fig. 1), and in culture the
cells can divide and differentiate into organized
tissue (Fig. 1). Upon separation from the cap,
metabolic activity in border cells increases and
gene expression undergoes a global switch, such
that mRNA and protein profiles are grossly distinct from those of progenitor cells in the root
cap5. Border cells then produce specific metabolites, such as anthocyanins and antibiotics6
(Fig. 1), and enzymes including a Rhizobiuminduced peroxidase7 and a low-pH galactosidase
(C. Jian and M.C. Hawes, unpublished; Fig. 1).
We call these unusual cells (previously termed
‘sloughed root cap cells’) root ‘border’ cells to
128
March 2000, Vol. 5, No. 3
emphasize that they constitute a biotic boundary layer between the root surface and the soil.
In the past, border cells largely have
been overlooked as a structural and functional
component of root systems. One reason for this
oversight is the notion that the cells are a moribund by-product of root-cap turnover. This presumption has prevailed in recent years in spite
“…environmental
signals
can override the control of
border cell development using
endogenous signals.”
of the fact that L. Knudson8 demonstrated that
it was incorrect in 1919 (Refs 1,2). Although
species in a few families, such as the
Compositae, have border cells that die before
detachment, in the majority of species examined, border cell populations are .90% viable5.
Border cells of maize and pea grown hydroponically can remain viable for more than three
months8, and cells of maize have been reported
to survive for a week or more in field soil9. A
second reason for the lack of attention paid to
border cells is that the cells disperse so efficiently in response to free water that even the
gentlest washing of root systems to remove soil
before experimental manipulation and examination removes all border cells1. However, in the
absence of free water the cells cling so tenaciously to the root periphery that their presence
is difficult to distinguish even by experienced
investigators.
Regulation of border cell production
The number of border cells that can be produced daily by a given root is conserved at the
family level, and can range from a dozen for
tobacco to .10 000 for cotton and pine. The
process of border cell production for tap roots,
lateral and branch roots appears to be similar.
Therefore, for complex root systems with hundreds of branches, millions of border cells
would be deposited into the soil daily if their
production were a constitutively expressed
process, which was assumed to be the case for
many years10. However, recent studies have
established that border cell production in
species such as cereals, legumes and cotton, is
a tightly regulated process that is controlled by
endogenous and environmental signals. In the
laboratory, radicles synthesize a set number of
border cells within 24 h of emergence. As they
accumulate on the cap periphery, border cells
of pea release an extracellular suppressor that
specifically inhibits mitosis in the root-cap
meristem, without affecting mitosis in the
adjacent apical meristem that gives rise to root
growth11. Root-cap turnover ceases and no
new border cells are made, but the set of
~4000 cells that has been made remains
appressed to the periphery of the cap as the
root elongates. If border cells are removed, or
if the extracellular suppressor is diluted by
dipping the root into water, cell division in the
root cap meristem resumes within 5 min, and
remains high for 2 h. New cells can be collected from the cap periphery within 1 h after
removal of the old ones; after 24 h a new set
of 4000 cells is complete and no more cells are
made. This process, which can be synchronized experimentally, has been exploited to
identify two genes, psugt1 and rcpme1, which
play a role at both ends of border cell development – cell division in the cap meristem and
cell separation at the cap periphery, respectively3,12. When the expression of either gene
in transgenic roots is inhibited by antisense
mRNA mutagenesis, border cell development
is blocked. Such mutants are being used under
controlled conditions to dissect the molecular
and cellular mechanisms by which border cell
development is regulated.
How border cell development is regulated
in soil is unknown. Because border cells do not
detach readily except when they are actually
placed into water, a root growing through conditions where the tip never experiences a flush
of water, could, theoretically, be expected to
retain the same set of border cells throughout
its lifetime. Renewed border cell production in
response to any exposure to free water, such as
during rain or irrigation, would be a logical
prediction. However, when roots are grown in
a wet semisolid matrix, such as water agar,
where border cells can be seen to be removed continuously when viewed microscopically, the results are not always found to be
1360 - 1385/00/$ – see front matter © 2000 Elsevier Science Ltd. All rights reserved. PII: S1360-1385(00)01556-9
trends in plant science
Perspectives
consistent with such straightforward predictions. Under these conditions, the extracellular suppressor, which inhibits cell division in
the root cap meristem, presumably would be
diluted continuously or would be removed by
physical forces as the tip elongates, thus a constant induction of mitosis with a continuous
production of border cells would be expected.
In some cases, continuous production does
occur, such that the entire root is ensheathed
in layers of detached border cells. In other
instances, sometimes in adjacent roots in a single Petri dish, border cell production occurs
only intermittently or not at all, suggesting that
localized conditions around individual roots
might influence the process1. Recent studies
have confirmed that environmental signals
can override the control of border cell development using endogenous signals13. For
example, when pea is exposed to increased
carbon dioxide levels, the normal control of
border cell production is overridden, such that
instead of producing the usual set of 4000
cells, twice as many cells are made (Fig. 2).
By contrast, border cell development in alfalfa
is impervious to increased carbon dioxide.
The regulation of this process at multiple
levels, with species-dependent variation in
responses to external signals, is one of the
properties that led us to suggest that border
cells can act as cellular ‘goalies’ (players
assigned to protect the goal in various sports;
Webster’s Unabridged Dictionary) to ward
off threats to the apical meristem.
Proposed functions of root border cells
The root cap provides physical protection for
the apical meristem; the border cells and their
associated mucilage were long presumed to
provide lubrication for the passage of the root
cap. To date, the hypothesis that border cells
lubricate the root tip is unsupported by experimental data2,14,15. The fact that plants turn
border cell production on and off without
hampering their ability to penetrate solid
media, and that a few species do not appear to
undergo border cell production at all argues
against such a ‘housekeeping’ function1,2 . We
have proposed that border cells serve plants by
enabling roots to define their own ecology.
The best characterized properties of border
cells involve their specific recognition and
responsiveness to soilborne microflora1,2. In a
genotype-dependent manner, border cells can:
• Attract zoospores (Fig. 1).
• Synthesize defensive structures in response
to fungal attack16.
• Repel or bind pathogenic bacteria (Fig. 1).
• Control growth and gene expression in
symbiotic bacteria (Fig. 1).
Chemotactic attraction to border cells of soilgrown plants has been shown to facilitate root
infection by pathogenic bacteria, and such
recognition might create ‘biased rhizospheres’
(b)
(a)
(h)
(c)
(g)
(d)
(e)
(f)
Fig. 1. Properties of root border cells. (a) Border cells have dispersed away from the cotton root
tip after immersion in water for 30 sec, showing one days’ accumulation of border cells
(~10 000) surrounding the tip (scale bar 5 1 mm). (b) The viability of detached border cells is
revealed by the accumulation of a vital stain, fluorescein diacetate (scale bar 5 10 mm). (c)
Cell division in border cells – induced by incubation in tissue culture medium containing plant
hormones28 (scale bar 5 10 mm). (d) Expression of the red-pigmented antibiotic shikonin is
localized in root border cells of Lithospermum erythrorhizon6 (scale bar 5 1 mm). (e)
Expression of a low-pH galactosidase is specific to root border cells of pea (C. Jian and M.C.
Hawes, unpublished) (scale bar 5 1 mm). (f) Border cells of cotton specifically attract
zoospores of Pythium dissotocum, which germinate, penetrate and kill the cells within 2 min
(Ref. 29) (scale bar 5 10 mm). (g) Genotype-specific binding of Agrobacterium tumefaciens
by border cells is correlated with susceptibility to infection19 (scale bar 5 10 mm). (h) When
pea roots are incubated on a lawn of Rhizobium leguminosarum expressing a nod-lacZ reporter
gene on solidified culture medium (pH 7.0), intense nod gene expression is detected in bacteria present at the root tip in the region where border cells are present30,31 (scale bar 5 1 mm).
by dictating which populations can colonize
plant root systems17. By modulating the properties of the external environment surround-
“…border cells can act as
cellular ‘goalies’…to ward off
threats to the apical meristem.”
ing the advancing root tip, the regulated
production of border cells could, directly or
indirectly, influence virtually all of the physical and chemical parameters (i.e. pH, nutrient
solubility or soil structure) of the root–
soil ecosystem. Here we focus on the direct
role that border cells might play in defense.
The recently discovered cellular responses of
border cells to nematodes, aluminum and
fungi suggest novel mechanisms by which
border cells might function to protect the root
tip from biotic and abiotic stress.
Border cells attract and immobilize
nematodes
In Greek mythology, the Sirens’ beautiful
songs tempted sailors to linger on their island,
and thus distracted them forever from completing their intended journeys. Recent studies
suggest that border cells might carry out a
similar distraction to pathogenic nematodes on
their way to infect root tips (X. Zhao and M.C.
Hawes, unpublished). When a pea root, whose
border cells have been removed, is placed onto
water agar containing a lawn of root knot nematodes, no sign of recognition is evident at the
root tip (Fig. 3). By contrast, when border cells
are present nematodes rapidly accumulate
around the root tip periphery (Fig. 3). When
March 2000, Vol. 5, No. 3
129
trends in plant science
Perspectives
are consistent with the hypothesis that border
cells have the capacity to synthesize an inducible extracellular structure that interferes
with the ability of aluminum to cause further
cellular damage. The mechanism by which this
occurs is unknown, but one possibility is that
the charged aluminum molecule becomes
immobilized by binding to the polysaccharide
layer20. If this is the case, the capacity of thousands of detached border cells to immobilize
aluminum ions could have a profound effect on
the ability of aluminium to damage the root tip,
the primary site of aluminum’s toxic effects21.
Border cells act as a decoy for
fungal infection
Fig. 2. Stimulation of border cell production by carbon dioxide. (a) In an ambient atmosphere,
roots of germinating seedlings produce ~4000 border cells in 24 h, cell production then ceases
unless the existing cells are removed. (b) In atmospheres containing increased carbon dioxide,
border cell production does not level off, and more than twice as many cells accumulate13.
these roots are lifted from the surface, clumps
of border cells left behind reveal high populations of actively motile nematodes (Fig. 3).
However, within 30 min, the motility of the
nematodes ceases and the worms appear
straight and inert, as they do when dead (Fig. 3).
When the immobilization phenomenon is measured using quantitative assays, a .99% loss of
mobility develops in response to a single heat-
stable, polar fraction of the root exudates (soluble extracellular material). This immobilization effect is reversible – within a few hours to
a few days (depending on conditions of the
assay) the nematodes resume full mobility. If
a similar process occurs in the soil, by the time
that the nematodes resume motility, a root
tip growing at a rate of 1 mm per hour
would be long past being in danger of penetration. This phenomenon might account, in
part, for the fact that the normal site of infection
by nematodes is behind the root tip, just past the
region where border cells are released18.
When pea seedlings are inoculated uniformly
with spores of pathogenic fungi, nearly all
develop visible lesions in the region just behind
the root tip within 24 h (U. Gunawardena and
M.C. Hawes, unpublished). A small proportion of seedlings also develop a visible lesion
at the root tip. When excised and placed onto
culture medium, fungal mycelium emerges
from these infected tips, confirming that the
root tip is not inherently resistant to infection
(Fig. 5). In the majority of seedlings (.90%),
the root tip remains white, and to the naked
eye appears to be infection free. However,
microscopic observation reveals that the tip is
actually covered in fungal hyphae (Fig. 5).
Mucilage production can repel bacteria
and decrease sensitivity to aluminum
Fig. 3. Attraction and immobilization of
the root knot nematode by root border
cells. (a) When a pea root is placed onto
water agar inoculated with nematodes,
nematodes are rapidly attracted to surround the border cells (right), but do not
surround roots whose border cells have
been removed (left). (b). When roots are
lifted from the surface, the active motility
of nematodes can be detected within
clumps of border cells left behind. (c)
After 30 min, motility within the clumps
of border cells ceases, and the nematodes
are rigid and inert.
130
March 2000, Vol. 5, No. 3
Border cells of legumes and cereals produce a
mucilage layer in response to co-cultivation
with pathogenic bacteria, but not in response
to E. coli19. This layer appears to repel the bacteria (Fig. 4), but its mechanism of action is
unknown. Recent studies have revealed that
co-cultivation of border cells with aluminum
results in the production of a mucilage layer
similar to that which occurs in response to
pathogenic bacteria (S. Miyasaka and M.C.
Hawes, unpublished). Within 2 h of exposure
to aluminum, a layer of mucilage around individual border cells increases in a dosagedependent manner, from being barely
detectable (Fig. 4), to being as wide (Fig. 4) or
wider than the cell’s diameter (Fig. 4).
Interestingly, the development of the layer is
correlated with a near-total cessation of aluminum-induced border cell death. For the first
2 h of exposure to aluminum, a dosage-dependent linear-rate of border cell death proceeds.
After 2 h, when the mucilage layer is in place,
the rate of cell death drops precipitously, such
that the viability of the cell population remains
largely static for the next 20 h (S. Miyasaka
and M.C. Hawes, unpublished). The results
Fig. 4. Development of a mucilage
layer by individual border cells. (a) A border cell of oats incubated for 2 h with
Agrobacterium tumefaciens has developed a mucilage layer that excludes the
bacteria, which can be seen at high density
outside the layer. (b) A border cell incubated in water and placed into India ink (a
stain that is excluded by polysaccharide)
exhibits a narrow layer of mucilage
around its surface. Border cells incubated
for 2 h in (c) 100 mM aluminum or in (d)
200 mM aluminum develop layers as wide
as or wider than the diameter of the cell.
trends in plant science
Perspectives
tear-away jerseys that were designed to detach
when grabbed by would-be tacklers, leaving
the runner free to keep moving down the field.
A similar mechanism, mediated by border
cell detachment, could, in part, account for
the observation that root tips generally escape
infection and colonization in the field2.
Model
Fig. 5. Mantle production in response to
pathogenic fungi (U. Gunawardena et al.,
unpublished). When roots are inoculated
with spores of Nectria haematococca,
most root tips remain white and appear
to be uninfected, but a few seedlings
develop visible lesions at the tip. When
such infected tips are plated onto culture
medium, abundant fungal growth
emerges (a, left). Most root tips remain
white and appear to be uninfected, but
when examined microscopically, are seen
to be covered in a sheath or ‘mantle’ of
fungal hyphae (b). When these roots are
placed into water and agitated gently, the
mantle of fungus falls away from the
root tip. When the tip is excised and
placed onto culture medium, no fungal
growth emerges (a, center), indicating
that the tips are sterile like the tips of the
uninoculated control seedlings (a, right).
When the tip is placed into water and agitated
gently, this ‘mantle’ detaches and can be seen
microscopically to consist of thousands of
digested border cells held together by actively
growing fungal mycelium2. Even under extreme
conditions, which include high concentrations
of spores and heat-shock temperatures, the root
tip remains sterile for days (Fig. 5), indistinguishable from uninoculated controls (Fig. 5) in
spite of having been covered by a sheath of
necrogenic fungus. In these plants, root growth
and development are comparable to that of the
controls, indicating that the apical meristems are
fully functional. In the 1950s, running backs
(American football players) were outfitted with
Taken together, the observed responses of border cells to aluminum, nematodes and pathogenic fungi are consistent with a twostep model in which the root tip is protected
by an unusual and complex process. First, the
root cap meristem actively produces a ‘front’
consisting of thousands of border cells that can
engineer the environment to render a threat
temporarily harmless. Elongating cells generated by the apical meristem then push the root
tip into and through the newly ‘engineered’
environment. The particular mechanism by
which the threat is rendered harmless by the
border cells might vary and could be constitutive (i.e. the secretion of a chemical that anesthetizes nematodes) or inducible in response
to a particular signal (i.e. the production of a
mucilage layer that immobilizes aluminum).
It is important to note again that only in the
presence of a flush of free water, such as after
rain or irrigation, would the en masse release
“…the capacity of thousands
of detached border cells to
immobilize aluminum ions
could have a profound effect
on the ability of aluminium
to damage the root tip, the
primary site of aluminum’s
toxic effects.”
of border cell populations occur. In the
absence of water, existing border cells remain
appressed to the root periphery and are carried
along as the root grows. It is also only in the
presence of free water that microorganisms or
toxins, such as aluminum, constitute a threat
to plants, because only then can movement
and uptake occur.
Reasons for protecting the tip in a
different way
The root tip is essential to plant health
This costly defensive process might be needed
because of the multifaceted role played by tissues within the root tip. One of the most-cited
functional distinctions between animals and
plants is that animals can move and plants cannot. Whereas animals can respond to signals
to avoid danger and seek nutrients, it is said
that plants must rely on internal signals to
Fig. 6. Directed movement of roots in
response to exposure to aluminum. Pea
seedlings with 1-cm-long roots were
placed onto medium containing no aluminum (left), 100 mM (center) or 200 mM
aluminum (right). Arrows indicate the
position of the root tip at the time that the
roots were placed onto the medium. In
response to aluminum, roots immediately began to grow away from the
plate at a 458 angle (S. Miyasaka and
M.C. Hawes, unpublished).
generate chemical responses in situ to synthesize their own food and protect themselves
from stresses. This simple dogma does not
take into account that complex root systems,
with each root tip elongating at a rate of 1 mm
per hour, can traverse .6 m of soil in a day22.
As in animals, this movement occurs as a
result of specific signals perceived by root
cells and transduced into behavioral responses
via common gene expression systems. Plant
roots move towards nutrients and water, and
away from toxins, such as aluminum (Fig. 6),
by sensing signals that notify them of the
presence of such stimuli. Anyone who has
ever battled to keep willow (Salix) roots out of
their septic tanks will appreciate the effectiveness of roots’ directed-movement capacities. Signals generated by the root as a result
of the detection of toxins or the uptake of
nutrients, are transmitted to the aboveground
parts of the plant. This communication
forms the basis, not only for the daily metabolism of the plant and its ultimate reproductive activities, but also for the overall size and
architecture of the whole plant23.
The primary site for perception of underground signals is the root tip23. Cells within the
terminal few millimeters of the root respond
to light, electricity, water, touch, ions and
other soluble molecules, to generate impulses
that direct the rate, direction and magnitude of
root movement. Just like animals, a capacity
to move away from danger and towards water
and other required elements is a crucial
component of plant life. So perhaps a more
important distinction between plants and animals is the mechanism by which movement is
generated: whereas animals can get up and
walk, plants can move only by the generation
and elongation of new cells by root meristems.
March 2000, Vol. 5, No. 3
131
trends in plant science
Perspectives
This newly synthesized tissue comprises the
same terminal few millimeters of tissue that
detects and responds to environmental stimuli
to condition movement toward nutrients and
away from danger. Not surprisingly, such
newly synthesized tissue is also especially
vulnerable to abiotic and biotic disease. For
humans, this is the functional equivalent of
having to rely on newborn babies to explore
and search a predator-filled wilderness for
supplies to send back to base camp. This biological dilemma might help to account for the
fact that in most species the vulnerable but
essential root meristems are surrounded by an
army of border cell ‘foot soldiers’2.
Other defensive strategies might not work
for the tip
We have proposed that a fail-safe defense
mechanism is needed to protect the root tip
because of its importance to the overall health
of the plant. However, it is well established
that plants already have complex defense
pathways that are activated rapidly in response
to external signals. Within moments of recognition of biotic and abiotic stresses, plants can
alter their surface structures, generate toxic
moieties, and activate the expression of genes
for the synthesis of dozens of antimicrobial
proteins and other metabolites24. Why would
an elaborate alternative method involving the
separation of masses of detached cells be
needed by plants that already are well
equipped to defend their cells? One explanation is that defense responses in plants often
involve the so-called ‘hypersensitive reaction’, in which the rapid necrosis of a few plant
cells is thought to inhibit the spread of a
pathogen25. In most tissues, including the
region behind the root tip where most
root–microbe relationships are initiated, the
sacrifice of a few cells offers containment of
the infection with little or no deleterious effect
on plant health. However, in the root tip, there
is no margin for the loss of a few cells in the
interest of containing infection when those
few cells comprise the apical meristem.
Indeed, experiments have confirmed that even
a minor necrotic reaction in the terminal 2 mm
of the root tip is correlated with the rapid cessation of growth of that root (U. Gunawardena
and M.C. Hawes, unpublished).
The activation of defense pathways in
response to biotic and abiotic signals does not
always result in hypersensitive cell death26.
The production of antimicrobial chemicals
can occur systemically, without any association with cell death. However, even in
the absence of cell death the activation of
defense pathways is a costly metabolic
process. As with other such responses (e.g.
heat shock), plants cope with this cost by
down-regulating other pathways. One activity that is known to be virtually shut down in
132
March 2000, Vol. 5, No. 3
response to the activation of defense pathways is the expression of genes needed for
cell division27. In most tissues, this would not
present a problem, but in a meristem it does.
In a meristem that plays a central role in
allowing the root system to avoid potentially
lethal stresses, a transient inhibition of
cell division during exposure to pathogens or
toxins could be deadly.
Future perspectives
Our model is congruent with the known biology of root infection: nearly all pathogenic
and symbiotic relationships are initiated not
“…in the root tip, there is no
margin for the loss of a few
cells in the interest of containing infection when those
few cells comprise the apical
meristem.”
at the tip, but in the region of elongation, just
behind the tip where the production of border
cells does not occur. We suspect that the main
reason that the effect of border cells on root
infection has never been examined, is its
effectiveness: so few pathogens and symbionts have been reported to initiate infection
or colonization at the root tip that there has
been little reason to focus on its role in infection. The availability of plants with altered
border cell production means that the most
obvious prediction of our model can now be
tested. If the presence of border cells is the
reason root tips are largely impervious to
infection, then inhibiting their production
and/or release will result in a change in the
site and/or nature of infection. If this is the
case, the study of how border cells can defend
the tip so successfully might elucidate novel
defense strategies that can be applied to
tissues that do not come equipped with
their own cellular ‘goalies’. It might not be
feasible to program other tissues to produce
and separate populations of border cells.
However, genetic engineering approaches
can be used to harness the cellular and extracellular mechanisms by which border cells
neutralize dangerous molecules or organisms
and express them in tissues that are not as well
protected as the root tip.
Acknowledgements
The authors thank Merritt R. Nelson for
the ‘tear-away jersey’ model for border cell
function, Hans D. VanEtten for critical reading of the manuscript, and the Dept of Energy,
Division of Energy Biosciences and United
States Dept of Agriculture for funds supporting basic research on the functions of root
border cells.
References
1 Hawes, M.C. and Brigham, L.B. (1992) Impact
of root border cells on microbial populations in
the rhizosphere. Adv. Plant Pathol. 8, 119–148
2 Hawes, M.C. et al. (1998) Function of root
border cells in plant health: pioneers in the
rhizosphere. Annu. Rev. Phytopathol. 36, 311–327
3 Wen, F. et al. (1999) Effect of pectin
methylesterase gene expression on pea root
development. Plant Cell 11, 1129–1140
4 Feldman, L.J. (1984) Development and dynamics
of the root apical meristem. Am. J. Bot. 7,
1308–1314
5 Brigham, L.A. et al. (1995) Differential
expression of proteins and mRNAs from border
cells and root tips of pea. Plant Physiol. 109,
457–463
6 Brigham, L.A. et al. (1999) Cell-specific
production and antimicrobial activity of
naphthoquinones in roots of Lithospermum
erythrothizon. Plant Physiol. 119, 417–428
7 Cook, D. et al. (1995) Transient induction of a
peroxidase gene in Medicago truncatula precedes
infection by Rhizobium meliloti. Plant Cell 7, 43–55
8 Knudson, L. (1919) Viability of detached root
cap cells. Am. J. Bot. 6, 309–310
9 Vermeer, J. and McCully, M.E. (1982) The
rhizosphere in Zea: new insight into its structure
and development. Planta 156, 45–61
10 Clowes, F.A.L. (1994) Origin of the epidermis in
root meristem. New Phytol. 127, 335–347
11 Brigham, L.A. et al. (1998) Meristem-specific
suppression of mitosis and a global switch in
gene expression in the root cap of pea by
endogenous signals. Plant Physiol. 118, 1223–1231
12 Woo, H.H. et al. (1999) Meristem-localized
inducible expression of a UDPglycosyltransferase gene is essential for plant
growth and development in pea and alfalfa. Plant
Cell 11, 2303–2315
13 Zhao, X. et al. (2000) Stimulation of border cell
production in response to increased CO2 levels.
Plant Physiol. 122, 181–188
14 Rogers, H.T. et al. (1942) The source and
phosphatase activity of exoenzyme systems of
corn and tomato roots. Soil Sci. 54, 353–365
15 Guinel, F.C. and McCully, M.E. (1986) Some
water-related physical properties of maize root
cap mucilage. Plant Cell Environ. 9, 657–666
16 Sherwood, R. (1987) Papilla formation in corn
root cap cells and leaves inoculated with
Colletotrichum graminicola. Phytopathology 77,
930–934
17 Hawes, M.C. and Smith, L. (1988) Requirement
for chemotaxis in pathogenicity of Agrobacterium
tumefaciens on roots of soil-grown pea plants.
J. Bacteriol. 171, 5668–5671
18 Prot, J.C. (1980) Migration of plant-parasitic
nematodes towards plant roots. Revue Nematol.
3, 305–318
19 Hawes, M.C. and Pueppke, S.G. (1987)
Correlation between binding of Agrobacterium
tumefaciens by root cap cells and susceptibility of
plants to crown gall. Plant Cell Rep. 6, 287–290
trends in plant science
Perspectives
20 Horst, W.J. et al. (1982) Mucilage protects root
meristem from aluminum injury.
Z. Pflanzenphysiol. Bd. 105, 435–444
21 Ryan, P.R. et al. (1995) Characterisation of
Al-stimulated efflux of malate from the apices of
Al-tolerant wheat roots. Planta 196,
103–110
22 Lynch, J. and van Beem, J.J. (1993) Growth and
architecture of seedling roots of common bean
genotypes. Crop Sci. 33, 1253–1257
23 Aiken, R.M. and Smucker, A.J.M. (1996) Root
system regulation of whole plant growth. Annu.
Rev. Phytopathol. 34, 325–346
24 Somssich, I.E. and Hahlbrock, K. (1998)
Pathogen defence in plants – a paradigm of
biological complexity. Trends Plant Sci. 3, 86–90
25 Kamoun, S. et al. (1999) Resistance to
oomycetes: a general role for the hypersensitive
response? Trends Plant Sci. 4, 196–200
26 Grant, M. and Mansfield, J. (1999) Early events
in host–pathogen interactions. Curr. Opin. Plant
Biol. 2, 312–319
27 Logemann, E. et al. (1995) Gene activation by
UV light, fungal elicitor or fungal infection in
Petroselinum crispum is correlated with
repression of cell cycle-related genes. Plant J. 8,
865–876
28 Hawes, M.C. and Pueppke, S.G. (1986) Sloughed
peripheral root cap cells: yield from different
species and callus formation from single cells.
Am. J. Bot. 73, 1466–1473
29 Goldberg, N. et al. (1989) Specific attraction to
and infection of cotton root cap cells by
zoospores of Pythium dissotocum. Can. J. Bot.
67, 1760 –1767
30 Zhu, Y. et al. (1997) Release of nodulation gene
inducing chemicals from root border cells. Plant
Physiol. 115, 1691–1698
31 Peters, N.K. and Long, S.R. (1988) Alfalfa
root exudates and compounds which promote
or inhibit induction of Rhizobium meliloti
nodulation genes. Plant Physiol. 88,
396–400
Martha C. Hawes*, Uvini Gunawardena
and Xiaowen Zhao are at the University
of Arizona, Dept of Plant Pathology,
Tucson, AZ 85721, USA;
Susan Miyasaka is at the University
of Hawaii, 461 W. Lanikaula St, Hilo,
HI 96720, USA.
*Author for correspondence (tel 11 520 621
5490; fax 11 20 621 9290;
e-mail mhawes@u.arizona.edu).
Errata
The review article by Richard A. Dixon and Christopher L. Steele (1999) Flavonoids and isoflavonoids – a gold mine for metabolic engineering. Trends Plant Sci. 4, 394–400 in the October 1999 issue of Trends in Plant Science contained two errors.
In Figure 2, dihydroflavonol (not flavanone) is acted on by flavonol synthase (FLS) to form flavonol. In Figure 4, the structure of the 2-hydroxyisoflavanone intermediate was depicted incorrectly. Corrected versions of these portions of the figures are
printed below.
OH
HO
OH
O
HO
OH
OH
(R1)
FLS
OH
OH
O
Flavanone
PII:S1360-1385(00)1572-7
H
(OH)
O
HO
O
HO
F3βH
O
Flavonol
O 2
O
OH
2-Hydroxyisoflavanone
OH
OH
O
Dihydroflavonol
Fig. 2.
Fig. 4.
Erratum
phyAr
phyAfr
Light
PII:S1360-1385(00)1573-9
phyBr
Fig. 1.
phyBfr
In the December 1999 issue of Trends in
Plant Science, we published a research
news article by Albrecht G. von Arnim
(1999) Phytochrome in the limelight.
Trends Plant Sci. 4, 465–466. In Figure 1,
top left it should read phyAr and not
phyBr. A corrected version of this portion
of the figure is printed left.
We apologize to the authors and to our
readers for these errors. A corrected version of these articles can be viewed in
HTML format at plants.trends.com
Students
Did you know that you are
entitled to a
50% discount on a
subscription to
Trends in Plant Science?
See the subscription order
card in this issue
for details.
March 2000, Vol. 5, No. 3
133