A comparative structural analysis of direct and indirect shoot ...

J. Plant Physiol. 157. 281–289 (2000)
© Urban & Fischer Verlag
http://www.urbanfischer.de/journals/jpp
A comparative structural analysis of direct and indirect shoot
regeneration of Papaver somniferum L. in vitro
Miroslav Ovečka 1 *, Milan Bobák 2 , Jozef Šamaj 3
1 Institute of Botany, Slovak Academy of Sciences, Dúbravská cesta 14, SK-842 23 Bratislava, Slovak Republic
2 Department of Plant Physiology, Faculty of Natural Sciences, Comenius University, Mlynská dolina B-2, SK-84215, Bratislava, Slovak Republic
3 Institute of Plant Genetics and Biotechnology, Slovak Academy of Sciences, Akademická 2, P.O. Box 39 A, SK-950 07 Nitra, Slovak Republic
Received January 11, 2000 · Accepted May 22, 2000
Summary
Cellular origin of shoot buds, cell morphogenesis and differentiation were studied during direct and
indirect shoot regeneration of Papaver somniferum L. in vitro. Direct shoot organogenesis was
induced in immature somatic embryos, where cell division and protomeristem formation started in
sub-epidermal and epidermal cell layers of hypocotyl. Indirect shoot regeneration was initiated from
callus culture using auxins and cytokinins, where compact globular meristemoids were produced.
The common morphogenetic event of direct and indirect shoot regeneration was an establishment of
non-random cell division and restricted cell expansion within the group of competent cells during
protomeristem formation. However, in contrast to direct regeneration, where all activated cells
became competent, in indirect regeneration, only peripheral cells of meristemoids acquired morphogenetic
competence. The second difference occurred in shoot tunica formation: original hypocotyl
epidermal cells participated in tunica formation during direct organogenesis, while this layer regenerated
de novo in meristemoids. These results indicate that cell morphogenesis during shoot regeneration
is independent of the developmental history of the competent cells.
Key words: morphogenesis – morphometry – Papaver somniferum L. – regeneration – shoot organogenesis
Introduction
Adventitious shoot regeneration is an important method of in
vitro plant biotechnology. Efficiency and yield of this type of
regeneration are closely related to culture initiation in some
species, regardless of whether shoot organogenesis can be
induced directly without callus production. A very important
* E-mail corresponding author: botuove@savba.savba.sk
factor determining culture response to a given culture conditions
is the state of differentiation of the participating cells. In
this respect, cell polarity, regulation of cell division, cell expansion,
and cell differentiation are important parameters in the
effort to understand the process of cell determination during
the early stages of shoot organogenesis (Šamaj et al. 1997).
In P. somniferum L., only callus induction from isolated
hypocotyls and indirect shoot organogenesis have been
achieved. A general rule of the induction was an application
0176-1617/00/157/03-281 $ 15.00/0

282 Miroslav Ovečka, Milan Bobák, Jozef Šamaj
of auxins and cytokinins (Nessler and Mahlberg 1979, Kamo
et al. 1982, Yoshikawa and Furuya 1985, Griffing et al. 1989),
or a transformation of the callus tissue originating from hypocotyl
by Agrobacterium rhizogenes (Yoshimatsu and Shimomura
1992). Organogenesis in the callus is connected to the
formation of meristematic centres–meristemoids with non-random
distribution of starch and lipids (Nessler and Mahlberg
1979, Šamaj et al. 1990 a, Ovečka et al. 1997). Both starch
and lipids have been shown to be related to shoot regeneration:
the amount of lipids in callus cells is 1.6–2.6 %, while
meristemoid cells contain 13 % in solid culture and 34.1% in
liquid culture (Yoshikawa and Furuya 1985). Cell specification
in meristemoids takes place during transition phase of organogenesis
(meristemoid “maturation”) when the initial accumulation
and subsequent utilisation of starch and lipids change
the nature of the cells. As a consequence, the meristemoid
cells express structural changes of nuclei (Bobák et al.
1990), plastids (Šamaj et al. 1988), and vacuolar system
(Šamaj et al. 1990 b). However, the precise data on cellular
origin of shoot bud primordia are scarce or nonexistent in
early works dealing with opium poppy shoot organogenesis.
In our previous study, we documented some cytological and
morphometrical differences among meristemoid cells during
indirect shoot organogenesis of P. somniferum L. (Ovečka et
al. 1997). Alternatively, bud production was achieved from
cultivated somatic embryos, and direct shoot regeneration of
opium poppy was documented for the first time (Ovečka et
al. 1997/98). The aim of the present work was to study the
similarities and differences in morphogenetical steps of shoot
primordia formation during direct and indirect shoot regeneration.
We characterised in detail: i. the cellular origin of
shoot primordia and shoot buds directly regenerated from
somatic embryos; and ii. the cell differentiation during shoot
primordia formation from meristemoids. In addition to cytological
and histological analyses, indirect shoot regeneration
was studied by morphometrical analysis.
Direct shoot regeneration was induced from somatic embryos of P.
somniferum L. Embryogenic callus culture was initiated from septa of
poppy capsules using a combination of α-naphtaleneacetic acid
(0.537–5.37 µmol/L) and kinetin (0.23–0.46 µmol/L). Regeneration of
somatic embryos took place on the growth regulator-free medium, as
described by Ovečka et al. (1996). Arrested torpedo somatic embryos
unable to develop functional root, spontaneously produced
secondary somatic embryos and adventitious shoots during cultivation
on the regeneration medium without growth regulators (Ovečka et
al. 1997/98). Organogenic callus culture was induced from unripe
seeds on solidified MS media (Murashige and Skoog 1962), supplemented
with 0.537 µmol/L α-naphtaleneacetic acid and 0.46 µmol/L
kinetin. Long-term shoot regeneration continued during culture cultivation
on the: (I) induction medium, (II) medium with 0.57 µmol/L indole-3-acetic
acid and 2.22 µmol/L 6-benzylaminopurine, and (III)
growth regulator-free medium (Ovečka et al. 1997).
Microscopy
Morphological observations were performed directly on living material
or using scanning electron microscopy (SEM). The samples for SEM
were fixed with 3 % glutaraldehyde (48 h, 0.1mol/L phosphate buffer,
pH 7.2) and 2 % OsO 4 (1h, the same buffer and pH). After washing in
buffer, the samples were dehydrated in ethanol, critical point dried in
liquid CO 2 , sputter coated with 20 nm layer of gold-palladium, and
observed with a JEOL JXA 840A (JEOL, Japan) microscope. Samples
for transmission electron microscopy (TEM) were fixed in 5 %
glutaraldehyde (5 h, 0.1mol/L phosphate buffer, pH 7) and 1% OsO 4
(2 h, 0.1 mol/L phosphate buffer, pH 7), dehydrated in acetone and
embedded in Durcupan ACM (Fluca, Buchs, Switzerland). Ultrathin
sections were stained according to Reynolds (1963) and observed in
ATEM 2000FX (JEOL, Japan) and TESLA BS 500 (Tesla, Czech Republic)
electron microscopes. Histological and cytological analyses
of shoot regeneration were performed at the light microscopy level.
Paraffin sections (7–10 µm) were prepared after sample fixation in FAA
(formalin 40 %, acetic acid 5 %, ethyl alkohol 50 %), embedded in
Histoplast S (Serva, Heidelberg, Germany) and stained with hematoxylin-eosin,
PAS reaction or Feuglen reaction. Sections (1–2 µm
thick) for the fine cytological observation were prepared from the
samples embedded for TEM. After SEM observation, some dried bulk
samples were immersed in ethanol and embedded in Histoplast S
(Serva, Heidelberg, Germany). The sections (7–10 µm thick) were observed
without additional staining under bright-field microscopy
(Ovečka and Bobák 1999). All prepared samples were studied using
ZEISS Jenalumar (Carl Zeiss, Jena, Germany) dissection light microscope.
Pictures digitalised using a Kappa CF 8/1 FMCC camera
(Kappa Messtechnik, Germany), and a Leica Q500MC image analysis
system (Leica Cambridge Ltd. England, UK) were processed
using Corel Photo Paint 8 (Corel Corporation, Canada).
Material and Methods
In vitro cultures
Morphometry
All planimetric measurements of meristemoid and shoot primordia
cells during indirect shoot regeneration were made using an ASBA
(Wild, Heerbrugg, Switzerland) image analyser. Cell size, cell shape,
and nucleus size were measured on the basis of cell cross-section
area in 1–2 µm thick sections. The mean values were compared
among the above-mentioned three culture media with different
amounts of growth regulators. The cell shape (form factor) was calculated
from the cell area and cell projections (a, b, a+b, a–b). The
form factor 5.09 means a circular shape, 6 means a square and 7 a
rectangle shape with a side ratio of 2 : 1 (Baluška et al. 1990). Morphometrical
values of cells in central and peripheral zones of the regenerated
shoot apical meristems were measured as standards for
the comparative characterisation of meristemoid cells. The computed
sizes, shapes of cells, and N/C ratios were used as parameters of
meristemoid cell activity (cell division and cell expansion). Linear regression
analyses, coefficients of correlation, and N/C ratios were calculated
using Sigma Plot (Jandel Scientific, USA) statistical software.

Shoot regeneration in opium poppy
283
Results
Shoot regeneration was initiated from hypocotyls of primary
somatic embryos (Fig. 1 a). The emerging shoot primordia
appeared on the hypocotyl surface of the somatic embryo
hypocotyl, but they did not disrupt the surface tissues (Fig.
1 b). In addition, no general tissue reorganisation was observed
(Fig. 1b). Due to these features, this kind of regeneration
was termed direct shoot organogenesis in this study.
Indirect shoot regeneration involved tissue dedifferentiation,
callus formation, and cell re-differentiation leading to de novo
shoot regeneration (Fig. 1 c). Tissue organisation and cell
morphogenesis of indirectly regenerated shoot primordia were
completely different in comparison to underlying callus tissue
(Fig.1d).
Direct shoot regeneration
Surface integrity of the embryo hypocotyl suggested that adventitious
shoots arose from superficial cell layers (Fig. 1 b).
Initially, sub-epidermal cells were activated and divided. The
plane of the first cell division seemed to be anticlinal (Fig.
2 a), but subsequent cell divisions were observed in both anti-
and periclinal planes (Fig. 2 b). Afterwards, dividing subepidermal
and epidermal cells created organogenic nodules,
the first recognisable structures on the hypocotyl surface
(Figs. 2 c, d). The two most important morphogenetic features
of these cells were dense non-vacuolated cytoplasm (Figs.
2 a, b, c) and reduced cell size (Fig. 2 d).
The first apparent indication of cell specification was observed
in the phase of apical meristem formation. The cells
within nodule apices adopted tunica-corpus zonation (Figs.
2 e, f). The original epidermal cells participated in differentiation
of the tunica layer (Fig. 2 c). This explains why the tunica
and protomeristem maintained morphological and histological
continuity between the forthcoming buds and the remaining
hypocotyl surface (Figs. 1b, 2 c, d, g). Cell specialisation
continued by both histodifferentiation of the shoot apical meristem
(Figs. 2 e, f) and transformation of organogenic nodules
into buds (Figs. 2 g, h). Adventitious buds were visible on
the hypocotyl surface (Figs. 2 g, h) and regular activity of the
apical meristem was detected, including formation and growth
of leaf primordia (Figs. 2 h, i).
Indirect shoot regeneration
Distinct mode of indirect induction of organogenesis was
based on callus production. Cell activation triggered a morphogenetic
switch in some cells within rarely dividing callus
tissue, resulting in the establishment of dividing, meristemlike,
and shoot-forming tissue (Fig. 3 a). Typical features of
meristemoids (representing population of competent cells)
were small cell size, cytoplasmic density, minimal vacuolation
(Figs. 3 a, b), and cell adhesion after cell division (Figs. 1 d,
3 a, b, c). The first cell differentiation events within the multicellular
meristemoids resulted in the formation of both peripheral
and central cell layers, with only peripheral cells continuing
their divisions (Fig. 3 b). Reserves were stored in central
cells in the form of starch (Figs. 3 b, d) and lipids (Figs. 3 e, f).
These cells were non-dividing but their cytological parameters
indicated high metabolic activity. Frequent endo- and
exocytosis in these cells (Fig. 3 e) and cell wall ingrowth for-
Figure 1. Shoot buds regenerated in the callus culture of P.
somniferum L. a. Somatic embryo (SE) in the culture where
the shoot (SH) regeneration from the hypocotyl was
induced. Bar = 1 mm. b. Directly regenerated shoots with
leaf primordia, visible as protuberances on the hypocotyl
surface of somatic embryo. Bar = 100 µm. c. Shoot bud
regenerated in the organogenic callus culture. The white
compact tissue represents meristemoids. Bar = 1 mm. d.
Shoot primordium on the surface of callus tissue. Bar =
100 µm.

284 Miroslav Ovečka, Milan Bobák, Jozef Šamaj
Figure 2. Direct shoot regeneration. a., b. Longitudinal section
of the embryo hypocotyl in the stage of cell activation.
a. First localised anticlinal cell divisions of sub-epidermal
cells (arrows). Bar = 50 µm b. Subsequent anticlinal and
periclinal cell divisions changed size and shape of sub-epidermal
activated cells. Bar = 50 µm. c. Localised cell divisions
in organogenic nodule. Frequent cell divisions of subepidermal
cells considerably reduced their size. First activated
epidermal cells are detected (arrows). Bar = 50 µm.
d. Organogenic nodules visible on the hypocotyl surface.
Bar = 50 µm. e. Concentration of cell proliferation in the
centre of organogenic nodule. Bar = 50 µm. f. Protomeristem
with established tunica-corpus zonation in the stage
of protomeristem transformation into shoot apical meristem.
Bar = 50 µm. g. Cross-section of the shoot bud primordium
in the vicinity of meristematic root pole of somatic embryo.
Note the involvement only of the epidermal and several
sub-epidermal cell layers in adventitious shoot production.
Bar = 50 µm. h. Surface view on the shoot apical meristem
with well defined three layered tunica (arrow) and initiation
of leaf primordia outgrowth (arrowheads). Bar = 100 µm. i.
Developing leaves from the flattened apical meristem of
adventitious shoot bud. Bar = 100 µm.
mations in the cells possessing numerous mitochondria (Fig.
3 g) indicated active cell-to-cell transport.
Division of the peripheral meristemoid cells was not properly
regulated. Divisions took place in all directions, resulting
in a random arrangement of meristemoid cells (Fig. 3 c).
However, the selection of dividing meristemoid periphery was
very important in cell determination during shoot primordia
formation. The change in morphogenesis was connected
with regulation of cell division of the outermost peripheral
meristemoid cells. The cells expanded in a radial direction
and divided periclinally (Fig. 4 a). Later, peripherally located
daughter cells formed tunica cell layer undergoing anticlinal
divisions (Fig. 4 b). Organogenesis by protomeristem formation
continued, regularly associated with starch depletion in
the shoot-forming meristemoids (Figs. 4 c, d). The shoot primordium
was established when cell proliferation in the procambium
region and cell division in the peripheral zone of
the apical meristem were detected (Figs. 4 c, e).
Cell division and cell differentiation in procambial region of
both directly and indirectly regenerated shoots were important
for differentiation of vascular tissue and laticifer system
typical for P. somniferum L. Laticifers appeared first as laticifer
initials (Fig. 4 f); later they formed an articulated system
during cell expansion (Fig. 4 g). Cell wall perforation also
contributed to laticifer coupling in the lateral direction to promote
anastomosing of laticifer arrays. Early stages of perforation
were characterised by thinner and elastic cell wall within
the site of future perforation, which allowed impressing and
partial movement of cell contents between neighbouring cells
(Figs. 4 h, i). After cell wall perforation had been completed,
typical multinuclear, articulated, anastomosing laticifer system
was observed in both developing shoot procambium
(Fig. 4 j) and developing leaves (Fig. 4 k) of adventitious
shoots regenerated in vitro.
Our histological observations revealed that peripheral meristemoid
cells represent the original site of shoot primordia
formation. We used morphometric measurements of undifferentiated
and dividing cells of regenerated shoot apical meristems
in our effort to properly characterise meristemoid
cells during the course of shoot primordia formation. All cells
in the central and peripheral zones of shoot apical meristems
expressed high correlation between the volumes of the cell
and nucleus (N/C ratio). When we compared the measured
N/C ratio values of central and peripheral cells within the
apical meristem of regenerated shoots and peripheral meristemoid
cells, we found a high degree of similarity (Fig. 5).
Cell size was only one distinct morphometrical parameter. As
the cells of peripheral and central zones of apical meristems,
peripheral meristemoid cells were larger. As the cells in the
rib-zone of shoot apical meristem (Fig. 6 a), central meristemoid
cells were larger. Mathematical evaluation of the cell
shape (form factor) revealed an additional difference: lower
values for peripheral meristemoid cells and higher values for
shoot apical meristem cells (Fig. 6 b). This means that periph-

Shoot regeneration in opium poppy
285
Figure 3. Indirect shoot regeneration. 1. Cell determination.
a. Production of meristemoids within the callus tissue. Bar =
50 µm. b. Tight cell adhesion and meristematic nature of
the meristemoid cells. Note the cytological differences
(cytoplasmic density, thickness and stainability of the cell
wall) between peripheral dividing cells and cells accumulating
starch granules. Bar = 50 µm. c. Numerous globular
meristemoids appearing mostly on the callus surface. Cell
arrangement in the meristemoids was random, but they had
compact, non-friable nature. Bar = 50 µm. d, e, f, g, – Ultrastructure
of the central meristemoid cells. d. Transparent
cytoplasm and large deposits of starch in the central cells.
Bar = 10 µm. e. Insertion of vesicular and matrix material
into the cell wall of central meristemoid cells by exocytosis.
L-lipid droplet. Bar = 2 µm. f. Lipid droplets which fill up the
cells in the transition layer between centre and periphery of
the meristemoid. Bar = 10 µm. g. Cell wall ingrowths (asterisk)
in the central meristemoid cell. Note the presence of
numerous mitochondria. Bar = 2 µm.
eral meristemoid cells were more rounded and cells of shoot
apical meristem were more oblong. However, unlike these differences
in cell size and shape, in both meristemoid periphery
(Fig. 7a) and shoot apical meristem (Fig. 7b), cell growth
was closely related to the similar negative correlation
between cell size and cell shape. Increase in the mean of cell
size was found to be followed by general decrease in the
mean of form factor (Figs. 7 a, b). In conclusion, most of the
cytological and morphometrical observations indicate that
only peripheral meristemoid cells can be identified as determined
cells during indirect shoot organogenesis.
Discussion
This study deals with cytological characterisation of de novo
shoot regeneration of P. somniferum L. in vitro, focusing from
the morphogenetical point of view on the differences between
direct and indirect shoot primordia formation. Shoot organogenesis
is a suitable experimental system for use in discerning
patterns of cell division and expansion in order to
compare direct and indirect caulogenetic determination.
Three developmental steps have been identified in distinct
shoot-forming cultures: (I) initiation of meristematic activity in
callus or explant as an expression of the morphogenetic
competence; (II) cell determination during formation of meristematic
nodules; and (III) differentiation of adventitious shoot
buds (Von Arnold and Eriksson 1985, Bobák et al. 1989,
Sharma and Bhojwani 1990, Colby et al. 1991, Lakshmanan
et al. 1997). In contrast to well established models of shoot
regeneration, the structural details of shoot primordia initiation
and development are missing in studies of organogenic
poppy callus culture. In addition, direct shoot organogenesis
of opium poppy has not been published before now. The
question here was what aspects of cell division, cell expansion,
and cell differentiation are related to cellular origin of
shoot primordia during either direct or indirect shoot regeneration.

286 Miroslav Ovečka, Milan Bobák, Jozef Šamaj
Figure 4. Indirect shoot regeneration. 2. Formation of shoot
primordia. a. Morphogenesis of peripheral meristemoid
cells during shoot primordia formation. Cells expanded in
anticlinal direction (arrowhead) and subsequently divided
periclinally (arrows). Bar = 50 µm. b. Establishment of the
tunica layer on the meristemoid surface by anticlinal cell
division. Bar = 50 µm. c. Histodifferentiation of the shoot
apical meristem (SAM). Bar = 50 µm. d. Cell proliferation
concentrated in the dome-like protomeristem. Cells located
away of the centre of cell division are devoid of starch and
start to vacuolate. Bar = 50 µm. e. Developed shoot bud
with leaf primordia and vascular tissue (VT) in the procambium.
Bar = 100 µm. f. Laticifer initials (arrows) in the procambial
region of shoot apical meristem. Bar = 50 µm. g.
Initial perforation of cross cell wall between two laticifer initials.
Bar = 50 µm. h. Formation of articulated, anastomosing
laticifer system. Note the elastic properties of longitudinal
cell wall in the site of future perforation (arrow). Bar =
50 µm. i. Insertions of cell volume between future anastomosing
laticifers in cross-section (arrows). Bar = 50 µm. j.
Articulated, anastomosed laticifer system in shoot procambium.
Bar = 50 µm. k. Articulated, anastomosed laticifer
system in developing leaves of regenerated shoots. Bar =
50 µm.
Figure 5. Similarity in the correlation curves between cell
size and nucleus size of peripheral meristemoid cells
(——) and cells of peripheral and central zones of the
apical meristems in regenerated shoots (——). 30–50
cells of each zone were measured on the medium supplemented
with 0.537 µmol/L of α-naphtaleneacetic acid and
0.46 µmol/L kinetin (I), or with 0.577 µmol/L indole-3-acetic
acid and 2.22 µmol/L 6-benzylaminopurine (II), or growth
regulator-free medium (III). Cell size values and nucleus
size values are presented as planimetric values of cell and
nucleus area measured from cross sections.
Direct shoot formation was initiated in the culture where
somatic embryos had failed to develop correctly, but maintained
continuously juvenile, undifferentiated state of their
cells in the root pole which were able to generate new somatic
embryos (Ovečka et al. 1997/98). However, as we show
here, adventitious shoots originated from the differentiated,
non-dividing tissue of hypocotyl. The competent cells have to
be re-activated and they enter cell division. The culture underwent
spontaneous repetitive regeneration including shoot
organogenesis and somatic embryogenesis (Ovečka et al.
1997/98). Some internal factors determining cell competence
in the course of shoot bud regeneration have been identified,
such as reduced cellulase activity in tobacco callus culture
(Truelsen and Ulvskov 1995), or importance of polar auxin
transport, because its inhibition by TIBA (2,3,5-triiodobenzoic
acid) strongly enhanced shoot regeneration, but not somatic

Shoot regeneration in opium poppy
287
Figure 6. Mean of the cell cross section area (a), and mean
of the form factor describing cell shape (b), observed during
shoot regeneration in the medium supplemented with
0.537 µmol/L of α-naphtaleneacetic acid and 0.46 µmol/L
kinetin. Peripheral meristemoid cells (A) were compared to
cells of central and peripheral zones of regenerated shoot
apical meristems (B), and central meristemoid cells (C)
were compared to cells in the rib-zone of shoot apical
meristems (D). Percentage amount of cells with isodiametric
(a side ratio app. 1 : 1), and oval shape (side ratio up to
2 : 1) is indicated (b). The rest of the cells had a more elongated
shape. Results from measurements of app. 300 cells
of each selected part of meristemoids and shoot apical
meristems. ** – significant at P

288 Miroslav Ovečka, Milan Bobák, Jozef Šamaj
Shoot regeneration in callus culture requires a combination
of auxins and cytokinins (Christianson and Warnick 1983, De
Klerk et al. 1997). The main change in cell division and cell
expansion was observed during competence acquisition.
Meristemoids were produced from rarely dividing callus cells.
Cell division was more patterned within meristemoids. Their
cells were small, micro-vacuolated, and able to stick together.
This indicates that the fate of the callus culture is altered
through cell morphogenesis of the activated competent
cells as a result of the action of growth regulators.
The second switching point in cell morphogenesis of
poppy callus culture occurred in the multicellular meristemoids
as a defined step of cell determination within the shootforming
tissue. A critical event was the localisation of cell
division to the meristemoid periphery. The cell determination
has been expressed when modified polar periclinal cell division
occurred in meristemoids. Apical daughter cells generated
dividing peripheral layers with morphogenetic competence
while the distal daughter cells became more differentiated
and generated storage cells possessing metabolic
sources. The previously observed differences in cell size and
shape showed that the peripheral cells were smaller, having
an oblong shape as a result of frequent cell division, while
the central cells became larger due to globular expansion
(Ovečka et al. 1997). Starch-containing cells were the largest
ones. They had isodiametric shape, and any changes (cell
division, plastid division, vacuolation) took place only after
starch utilisation had been initiated (not shown). Central
meristemoid cells thus can support organogenesis by the
energy. It was clearly shown that some metabolic indications
of shoot regeneration (protein inclusions, starch, lipids, high
enzyme activities) were found only in meristemoids (Ross et
al. 1973, Nessler and Mahlberg 1979, Patel and Berlyn 1983).
Meristemoids possessed a high starch content, while high
activities of starch-degrading enzymes were observed in
shoot buds (Thorpe and Murashige 1970, Thorpe and Meier
1974, Patel and Berlyn 1983). A high N/C ratio of peripheral
meristemoid cells of opium poppy, comparable with cells in
shoot apical meristem (Fig. 5), together with comparable rate
of chromatin de-condensation (Ovečka et al. 1997) are also
important parameters of cell determination within the meristemoids.
Variability in the size of nucleus and the correlation
with DNA amount is closely related to shoot regeneration
(Flinn et al. 1989, Fournier et al. 1991). A two-phase organogenesis,
with meristemoid formation during the first phase
and formation of globular structures from the surface during
the second phase, was observed in sundew callus tissue
(Bobák et al. 1993). This data indicates that “mitotic zonation”
established a distinct morphological competence of meristemoid
cells in the process of shoot primordia formation
when the site of origin of shoot buds was located at the
meristemoid periphery.
Concluding our comparative analysis of direct and indirect
shoot regeneration in P. somniferum L., important changes in
cell morphogenesis, characterised by non-random cell divisions,
took place during dermal layer and protomeristem formation
in shoot primordia. We observed a random orientation
of cell divisions in meristemoids, while protomeristem and especially
tunica formation were accompanied by regulated
periclinal and subsequently anticlinal cell divisions. In directly
regenerated shoots, the dermal layer was formed by original
epidermal cells through anticlinal cell divisions. This type of
division maintains epidermal cell polarity. Peripheral meristemoid
cells are polarised with non-homogenous cell surfaces
due to the presence of adhered and outer non-adhered cell
walls (Ovečka and Bobák 1999). This polarity and cell position
are important morphogenetic factors in cell determination.
Cell expansion was found to be minimal in cells involved
in formative, morphogenetically important cell division. The
cells expanded, but not over the cell volume necessary for a
subsequent cell division. In addition, cell size and cell shape
have to be negatively correlated. This relationship would indicate
an independence of post-mitotic cell growth and an extensive
cell vacuolation, originally observed in maize root
apex (Baluška et al. 1990). A comparative analysis revealed
that cell activation in shoot organogenesis of P. somniferum
L. depends on the stage of cell differentiation before initiation,
while cell morphogenesis of the determined cells is similarly
regulated both spatially and temporally, during direct
and indirect shoot regeneration.
Acknowledgements. We gratefully acknowledge D. Jantová for technical
assistance, and Dr. M. Čiamporová for the English correction.
This work was supported by Grant Agency VEGA, Grant No.1/6030/99
and Grant No. 1/6107/99 from the Slovak Academy of Sciences. M.O.
gratefully acknowledges the Alexander von Humboldt-Stiftung (Bonn,
Germany) for donation of the Leica Q500MC image analysis system.
References
Altamura MM, Capitani F, Falasca G, Gallelli A, Scaramagli S, Bueno
M, Torrigiany P, Bagni N (1995) Morphogenesis in cultured thin
layers and pith explants of tobacco. I. Effect of putrescine on cell
size, xylogenesis and meristemoid organization. J Plant Physiol
147: 101–106
Baluška F, Kubica Š, Hauskrecht M (1990) Postmitotic “isodiametric”
cell growth in the maize root apex. Planta 181: 269–274
Bobák M, Blehová A, Erdelský K (1989) Histological and cytological
study of early phases of organogenesis on excised leaves of Drosera
spathulata L. cultivated in vitro. Biológia (Bratislava) 44: 785–792
Bobák M, Blehová A, Šamaj J, Ovečka M, Krištín J (1993) Studies of
organogenesis from the callus culture of the sundew (Drosera
spathulata Labill.). J Plant Physiol 142: 251–253
Bobák M, Šamaj J, Erdelský K (1990) Ultrastructural polymorphysm of
interkinetic nuclei and nucleoli in the cells of organogenic callus
cultures of Papaver somniferum L. in in vitro conditions. Acta FRN
Univ Comen – Physiol Plant XXV: 3–20
Bronner R, Jeannin G, Hahne G (1994) Early cellular events during organogenesis
and somatic embryogenesis induced on immature
zygotic embryos of sunflower (Helianthus annuus). Can J Bot 72:
239–248

Shoot regeneration in opium poppy
289
Burritt DJ, Leung DWM (1996) Organogenesis in cultured petiole explants
of Begonia × erythrophylla: the timing and specificity of the
inductive stimuli. J Exp Bot 47: 557–567
Choi YE, Kim HS, Soh WY, Yang DC (1997) Developmental and structural
aspects of somatic embryos formed on medium containing
2,3,5-triiodobenzoic acid. Plant Cell Rep 16: 738–744
Christianson ML, Warnick DA (1983) Competence and determination
in the process of in vitro shoot organogenesis. Dev Biol 95: 288–
293
Colby SM, Juncosa AM, Stamp JA, Meredith CP (1991) Developmental
anatomy of direct shoot organogenesis from leaf petioles of
Vitis vinifera (Vitaceae). Amer J Bot 78: 260–269
De Klerk G-J, Arnholdt-Schmitt B, Lieberei R, Neumann K-H (1997)
Regeneration of roots, shoots and embryos: physiological, biochemical
and molecular aspects. Biol Plant 39: 53–66
Flinn BS, Webb DT, Newcomb W (1989) Morphometric analysis of reserve
substances and ultrastructural changes during caulogenic
determination and loss of competence of Eastern white pine (Pinus
strobus) cotyledons in vitro. Can J Bot 67: 779–789
Fournier D, Bonnelle Ch, Tourte Y (1991) Ultrastructural features of
soybean somatic cells at the beginning of an organogenic process:
toward a new concept. Biol Cell 73: 99–105
Gilissen LJW, Van Staveren MJ, Hakkert JC, Smulders MJM, Verhoeven
HA, Creemers-Molenaar J (1994) The competence of cells for cell
division and regeneration in tobacco explants depends on cellular
location, cell cycle phase and ploidy level. Plant Sci 103: 81–91
Griffing LR, Fowke LC, Constabel F (1989) Radioimmunoassay of
morphinan alkaloids in Papaver somniferum hypokotyls, callus and
suspension cultures. J Plant Physiol 134, 645–650
Kamo KK, Kimoto W, Hsu AF, Mahlberg PG, Bills DD (1982) Morphine
alkaloids in cultured tissues and redifferentiated organs of Papaver
somniferum. Phytochemistry 21: 219–222
Konieczny R (1996) Plant regeneration from immature embryo culture
of Trifolium michelianum Savi. – histological observations on adventitious
shoot induction. Acta Soc Bot Pol 65: 261–266
Lakshmanan P, Ng SK, Loh CS, Goh CJ (1997) Auxin, cytokinin and
ethylene differentially regulate specific developmental states associated
with shoot bud morphogenesis in leaf tissues of mangosteen
(Garcinia mangostana L.) cultured in vitro. Plant Cell Physiol
38: 59–64
Marcotrigiano M (1986) Origin of adventitious shoots regenerated
from cultured tobacco leaf tissue. Amer J Bot 73: 1541–1547
Murashige T, Skoog F (1962) A revised medium for rapid growth and
bioassay with tobacco tissue cultures. Physiol Plant 15: 473–497
Nessler CL, Mahlberg PG (1979) Ultrastructure of laticifers in redifferentiated
organs on callus from Papaver somniferum. Can J Bot 57:
675–685
Ovečka M, Bobák M (1999) Structural diversity of Papaver somniferum
L. cell surfaces in vitro depending on particular steps of plant
regeneration and morphogenetic program. Acta Physiol Plant 21:
117–126
Ovečka M, Bobák M, Erdelský K, Šamaj J, Blehová A, Krištín J (1996)
Morphology and conversion ability of somatic embryos in longterm
embryogenic callus culture of Papaver somniferum. Biologia
(Bratislava) 51: 417–423
Ovečka M, Bobák M, Šamaj J (1997) Development of shoot primordia
in tissue culture of Papaver somniferum L. Biol Plant 39: 499–506
Ovečka M, Bobák M, Blehová A, Krištín J (1997/98) Papaver somniferum
regeneration by somatic embryogenesis and shoot organogenesis.
Biol Plant 40: 321–328
Patel KR, Berlyn GP (1983) Cytochemical investigations on multiple
bud formation in tissue cultures of Pinus coulteri. Can J Bot 61:
575–585
Reynolds ES (1963) The use of the lead citrate at high pH as an electron
opaque in electron microscopy. J Cell Biol 17: 208–213
Ross MK, Thorpe TA, Costerton JW (1973) Ultrastructural aspects of
shoot initiation in tobacco callus cultures. Amer J Bot 60: 788–795
Sharma KK, Bhojwani SS (1990) Histological aspects of in vitro root
and shoot differentiation from cotyledon explants of Brassica juncea
(L.) Czern. Plant Sci 69: 207–214
Šamaj J, Bobák M, Erdelský K (1990a) Histological-anatomical studies
of the structure of the organogenic callus in Papaver somniferum
L. Biol Plant 32: 14–18
Šamaj J, Bobák M, Erdelský K (1990 b) Vacuolar system in the cells of
organogenic callus cultures of Papaver somniferum L. in vitro.
Acta FRN Univ Comen Physiol Plant XXV: 21–32
Šamaj J, Bobák M, Erdelský K (1988) Plastids in cells of Papaver
somniferum L. organogenic calli grown on various media. Biológia
(Bratislava) 43: 577–588, in Slovak
Šamaj J, Bobák M, Ovečka M, Blehová A, Pret’ová A (1997) Structural
features of plant morphogenesis in vitro. Veda, Bratislava
Tétu T, Sangwan RS, Sangwan-Norreel BS (1990) Direct somatic embryogenesis
and organogenesis in cultured immature zygotic embryos
of Pisum sativum L. J Plant Physiol 137: 102–109
Thorpe TA, Murashige T (1970) Some histochemical changes underlying
shoot initiation in tobacco callus cultures. Can J Bot 48: 277–
285
Thorpe TA, Meier DD (1974) Starch metabolism in shoot-forming tobacco
callus. J Exp Bot 25: 288–294
Truelsen TA, Ulvskov P (1995) The role of cellulase in hormonal regulation
of shoot morphogenesis in tobacco callus. Planta 196: 727–
731
Von Arnold S, Eriksson T (1985) Initial stages in the course of adventitious
bud formation on embryos of Picea abies. Physiol Plant 64:
41– 47
Yoshikawa T, Furuya T (1985) Morphinan alkaloid production by tissue
differentiated from cultured cells of Papaver somniferum. Planta
Med 44: 110–113
Yoshimatsu K, Shimomura K (1992) Transformation of opium poppy
(Papaver somniferum L.) with Agrobacterium rhizogenes MAFF
O3-O1724. Plant Cell Rep 11: 132–136

ACTA
PHYSIOLOGIAE
PLANTARUM
Vol. 21. No. 2. 1999:117-126
Structural diversity of Papaver somniferum L. cell surfaces in vitro depending
on particular steps of plant regeneration and morphogenetic program
Miroslav Overka, Milan Bobrk*
Institute of Botany, Slovak Academy of Sciences, Dfibravskfi cesta 14, SK-84223 Bratislava, Slovak Republic
*Department of Plant Physiology, Faculty of Natural Sciences, Comenius University, Mlynskfi dolina B-2,
SK-84215, Bratislava, Slovak Republic
Key words: cell surface, cell adhesion, extracellular
matrix, morphogenesis, Papaver somniferum
L., plant regeneration, somatic embryogenesis,
shoot organogenesis
Abstract
Culture of Papaver somniferum in vitro was used for a characterisation
of cell surface structures and mode of cell adhesion
and cell separation during cell differentiation and plant regeneration
in somatic embryogenesis and shoot organogenesis. In
early stages of somatic embryogenesis, cell type-specific and
developmentally regulated change of cell morphogenesis was
demonstrated. Cell wall of separated embryonic cells were
self-covered with external tubular network, whereas morphogenetic
co-ordination of adhered cells of somatic proembryos
was supported by fine and fibrillar external cell wall continuum
of peripheral cells, interconnecting also local sites of cell separation.
Such type of cell contacts disappeared during histogenesis,
when the protodermis formation took place. Tight cell
adhesion of activated cells with polar cell wall thickening, and
production of extent mucilage on the periphery were the crucial
aspects of meristemoids. Fine amorphous layer covered developing
shoot primordia, but we have not observed such comparable
external fibrillar network. On the contrary intercellular
separation of differentiated cells in regenerated organs, and accepting
distinct develdpmental system of somatic embryogenesis
and shoot organogenesis, cell adhesion in early stages
and ultrastructural changes associated with tissue disorganisation,
and the subsequent reorganisation into either embryos or
shoots appear to be regulatory morphogenetical events of plant
regeneration in vitro.
Introduction
Cell and tissue cultivation in vitro induces activation
of different cellular response mechanisms, mediating
cell adaptation under new environmental
conditions. As a consequence of cell adaptation, the
shift of existing morphogenetic program by inducing
of undifferentiated cell state, and (or) reactivation
of cells in a new redifferentiation program
could be manipulated by in vitro conditions. Cell
reactivation during induction of somatic embryogenesis
is connected with the reprogramming of
gene expression and reorganisation of internal cellular
architecture, affecting cell morphology and
pattern of cell division (Cyr et al. 1987, Dijak and
Simmonds 1988, Dudits et al. 1991, Emons 1994).
Cell morphology and cell synchrony during plant
regeneration in vitro are controlled by particular
morphogenesis, where cell-to-cell communications
are responsible for recognition of cell Position
and behaviour in early stages of regeneration. Cell
position with a population of meristematic cells appears
to be an important factor in the determination
of cell fate (Poethig 1989). Intercellular communications
together with perception and transduction
of environmental stimuli could be of different nature.
From the morphological and physiological
117

M. OVECKA & M. BOBf~K
point of view redefined plant cell wall (Roberts
1989) appears to be a dynamic and active cellular
compartment, mediating cell signalling through the
special wall integral part referred to as an extracellular
matrix (Roberts 1994). It is not surprising that
plant cell and tissue surfaces became very attractive
objects in the study of cell adhesion and separation
during the cell differentiation and tissue morpho-
~-enesis (Roberts 1989, 1994, Knox 1992a, b).
In embryogenic culture of Papaver somniferum L.
in vitro, we evaluated morphological variability of
somatic embryos (Ove~ka et al. 1996), and the secondary
regeneration ability of embryo cells during
subsequent long-term cultivation was described
(Ove~ka et al. 1997/98). The aim of this study is the
description of the fine structure of cell and early
embryo surfaces (including cell wall, and structures
located extracellularly) during long-term somatic
embryogenesis, compared to the structural characteristics
of cell and tissue surface organisation in
early stages of opium poppy shoot regeneration in
vitro.
Materials and Methods
Somatic embryogenesis of Papaver somniferum L.
was initiated and maintained as previously described
(Ove~ka et al. 1996). Briefly, embryogenic
callus culture was induced from unripped seeds on
MS induction medium (Murashige and Skoog
1962), supplemented with various concentrations
of ct-naphtaleneacetic acid and kinetin. Somatic
embryos regenerating on hormone-free medium
followed a long-term proliferation with the capacity
of secondary somatic embryogenesis.
Long-term organogenic culture of Papaver somniferum
L. was induced from unripe seeds on Murashige
and Skoogs (1962) medium, supplemented
with c~-naphtaleneacetic acid and benzylaminopurine,
or indoleneacetic acid and kinetin (Samaj et al.
1990, Ove6ka et al. 1997).
Resin-embedded samples for transmission electron
microscopy (TEM) were fixed in 5 % glutaraldehyde,
buffered with phosphate buffer for 5 h, and
postfixed in 2 % osmium tetroxide for 2 h, buffered
with the same buffer. After dehydration in acetone,
the samples were embedded in Durcupan ACM
(Fluca). Ultrathin sections stained with uranyl acetate
and lead citrate were examined using Tesla BS
500 electron microscope. Semithin resin sections of
1-1.5 pm thickness prepared for light microscopy
were stained with 1% aqueous toluidine blue and 2
% aqueous basic fuchsine. Samples for scanning
electron microscope (SEM) were fixed in 3 % buffered
(phosphate buffer) glutaraldehyde for 48 h and
2 % buffered osmium tetroxide for 1 h. Samples dehydrated
in ethyl alcohol were critical point dried in
CO2, sputter-coated with gold (20 nm) and examined
in scanning electron microscope JXA 840A
(JEOL) at 15 kV. Some critical point dried samples
were immersed in ethyl alcohol, transferred to butyl
alcohol, infiltrated and embedded in Histoplast S
(Serva). Sections of 8-10 pm thickness were dewaxed
and prepared for light microscopy without
additional staining, or stained with the periodic
acid-Shifts reaction (PAS).
Results
Somatic embryogenesis was induced in embryogenic
callus culture by activation and determination
of competent cells with growth regulators, and
competent cells expressed their embryogenic potential
on hormone-free medium. Cetlular organisation
and cell adhesion were different depending
on particular steps of embryonic activation, determination
and expression. Population of competent
cells were arranged in compact embryogenic clusters,
located predominantly in surface and subsurface
cell layers of callus. The cells in clusters undergoing
frequent cell divisions maintained
meristemic-like appearance, however random orientation
of cell division plane determined cell rearrangement
within the clumps after cell division
(Fig. 1 a). On the other hand, differentiated protodermis
of globular, heart-shaped and torpedo somatic
embryos comprising files of tightly arranged,
convex-shaped cells elongated in the shoot-root direction
was the most remarkable aspect of surface
morphology of regenerated somatic embryos since
the late globular stage (Fig. 1 b). Embryogenic clusters
included somatic proembryos arising from
competent cells together with later developmental
stages of somatic embryos (Fig. lb). When secondary
somatic embryogenesis was induced from the
primary somatic embryos, similar clusters of differ-
118

CELL SURFACE STRUCTURE AND MORPHOGENESIS IN VITRO
Fig. 1
119

o
"i
J

CELL SURFACE STRUCTURE AND MORPHOGENESIS IN VITRO
Fig. 3
Fig. 4
121

M. OVECKA & M. BOB,4K
Fig. 1. Scanning electron microscopy of critical point dried embryogenic tissue culture.
a. Dividing activated ceils and somatic proembryos in clusters on the surface of embryogenic callus culture. Bar = 100 lain
b. Asynchronously developing somatic proembryos (p), globular (g), heart-shaped (h) and torpedo (t) somatic embryos. Bar = 100
lam
c. Clustered secondary somatic embryos (arrows) originating from primary somatic embryo (E). Bar = 500 ~m
d. External interconnecting strip-like and fibrillar "bridges" among individual proembryos within the cluster in early stages of regeneration.
Bar = 10 ~am
e. Surface cell wails were covered with continuous discrete layer, which partly disappeared into fibrillar network in the sites of cell
separation (arrowheads). Bar = 10 lam
f. Distinct mode of cell adhesion and cell separation between embryonic competent cells (A) and protodermal cells of globular somatic
embryo (B). Embryonic cells were self-covered with tubular extension of the cell wall, which were absent on the surface of
the protodermis. Bar = 10 IJm
g. Smooth protodermis of heart-shaped somatic embryo. Bar = 10 lam
Fig. 2. Organogenic culture.
a. Globular meristemoid shape supported by anticlinal (thick arrow) and periclinat (thin arrow) cell divisions of isodiametric cells.
Mucilage layer (asterisk). Bar = 50 lam
b. Transmission electron microscopy of peripheral meristemoid ceils. Note the apparently thicker outer "boundary" cell walls of the
meristemoid with presence of wall-connecting elemental fibrils (arrows). Bar = 5 IJm
c. Initial stages of globular meristemoid formation (arrows). Bar = 100 lam
d. Synchrony in cell death of the senescing meristemoid. Bar = 100 Inn
e. Two meristemoids surrounded by the dense amorphous mucilage. Bar = 50 lam
f. Extent mucilage accumulation in the region of meristemoid senescence. Bar -- 100 lam
g. Giant starch grains as a source of metabolic energy for future organogenesis in promeristemoid cells. Bar = 50 lam
Fig. 3. Scanning electron microscopy of shoot regeneration.
a. Developing shoot primordium (asterisk) between callus cells. Note the presence of fine layer of amorphous material on the cell
surface (arrows). Bar = 100 lain
b. Leaf of regenerated shoot. Bar = 100 lam
c. Smooth surface of leaf epidermal and guard stomatal cells. Bar = I0 lam
Fig. 4. Formation of intercellular spaces.
a. Intercellular reticular structure in developing leaf tissue. Bar = 1 lam
b. Reticular (R) and fibrillar (F) filling of intercellular space within the developed leaf parenchyma. Ch-chloroplast, V-central vacuole.
Bar = 1 lam
c, d. Light microscopy of serial section of critical point dried parenchymatic tissue in the region of vascular strands of primary somatic
embryo during secondary somatic embryogenesis. Extracellular fibrillar and reticular network was present in large intercellular
spaces. The majority of PAS reaction-positive material occurred on the surface of cell walls (arrows). Bar = 50 lam.
entiating somatic embryos were produced (Fig. lc).
Detailed surface view shows fine structures covering
individual proembryos, connecting them to
each other (Fig. ld). Strip-like and fibrillar strands
bridged proembryos over distances among them
continuing around the neighbouring "connected"
proembryos. Strip-like bridges interconnected proembryos
over longer distances and mostly fine fibrillar
network covered tight, short distances (Fig.
ld). Similar fibrillar network was found to interconnect
surface proembryo ceils themselves continuing
as a more or less coherent layer on the cell
surface (Fig. le). The fine network could cover the
proembryo itself probably since the first cell divisions
(not shown), but single cells in embryogenic
cluster were covered by long tubular extensions radiating
from the surface of each globular cell (Fig.
lf). Late globular somatic embryos where the pro-
todermis have been established together with
heart-shaped and torpedo somatic embryos with
differentiated protodermis (Fig. lg) lacked completely
tubular or fibrillar extensions of outer tangential
cell walls (Figs. le, f) including fibrillar and
strip-like connections, which were observed among
the proembryos (Fig. ld).
Formation of globular meristemoids as a consequence
of frequent mitotic division of activated
cells took place during shoot organogenesis. Meristemoid
cells achieved and propagated their globular
appearance through periclinal, anticlinal and mixoriented
cell divisions predominantly in the meristemoid
periphery followed by isotropic cell expansion
(Fig. 2a). Fibrillar network was not observed
on the meristemoid surface (not shown), however
peripheral cells expressed position-dependent
assymetry in cell wall thickness. Thicker outer cell
122

CELL SURFACE STRUCTURE AND MORPHOGENESIS IN VITRO
walls (Fig. 2b) serve as a mechanical meristemoid
separation apart from the surrounding tissue and
conditions localised outside. Globular proliferation
of the meristemoids was clearly visible indicating
division synchrony since early stages of development
(Fig. 2c). Globular shape (or shape deformation
depending on the inter-meristemoid pressure)
thicker outer periclinal wall of peripheral cells and
cell synchronisation persisted throughout the meristemoid
development until the meristemoid life
terminated in the case of unefficient organogenesis
without formation of the shoot primordium (Fig.
2d). Meristemoid surfaces and relating callus cells
were covered by mucilaginous layer extently
stained with basic fuchsine (Fig. 2e). This extracellularly
located mucilage was present around the
meristemoids during their differentiation (Fig. 2a)
where the first indications of mucilage accumulation
on the meristemoid surface appeared in the
stage of frequent cell division of peripheral cells
(Fig. 2b). Mucilage accumulation was supported by
initiation of starch utilisation in the meristemoids
and the amount of external mucilage increased during
the long-term culture cultivation presumably in
the regions of meristemoid senescence (Fig. 2f).
Development of the meristemoids was closely related
to sugar metabolism and restricted to shootforming
tissue. Extent local starch accumulation
preceded formation of meristemoids, when the surrounding
mucilage was absent (Fig. 2c, g). Formation
of shoot primordia from meristemoids required
utilisation of starch reserves (not shown) and decreasing
starch amount in some meristemoids (Fig.
2e) or complete starch depletion in collapsed meristemoids
(Fig. 2d, f) correlated with increasing
amount of mucilage on their surface (Figs. 2d, e, f).
Developing shoot primordium was free of fibrillar
matrix on the surface (Fig. 3a), however underlying
shoot-forming meristemoids were overlaid by fine
layer of amorphous material (Fig. 3a). When shoots
were regenerated the outgrowing leaves contained
discrete epidermis (Fig. 3b) with developed stomata
(Fig. 3c) but lacking both fibrillar and amorphous
surface structures (Fig. 3c). Other most conspicuous
extracellular material appeared in intercellular
spaces of organised parenchyma tissues of
developed leaves in the organogenic culture and in
the tissues of primary somatic embryos. Intercellular
spaces of leaf chlorenchyma tissue possessed
more or less dense fibrilIar or reticular network
(Figs. 4a, b). Internal histological tissue organization
of primary somatic embryos became malformed
during secondary somatic embryogenesis
forming large intercellular "caves". These intercellular
spaces were filled with a dense reticular network
radiating from the surface of the parenchyma
cells (Figs. 4c, d). The type of extracellular matrix
interconnecting cells of re-differentiated tissues
and organs expressed different structural characteristics
compared to external cell and tissue surfaces
in early stages of plant regeneration.
Discussion
Histological and morphological study of opium
poppy culture in vitro revealed distinct modes of
cell adhesion, cell-to-cell communication and cell
surface structures in different stages of plant regeneration.
In somatic embryogenesis, the relationship
between cell shape, cell connection and presence or
absence of cell wall external fibrillar network in
embryonic tissue and differentiating somatic embryos
was clearly demonstrated. Competent embryonic
cells were localized in dividing regions on
the surface of the long-term cultivated embryogenic
callus tissue. Small cell size and mostly isodiametric
cell shape show their meristematic character.
Their cell walls expressed tubular extensions
on the surface of the embryogenic clusters. Somatic
proembryos formed in the embryogenic clusters
were composed of similar isodiametric cells but
fibrillar network interconnected all peripheral proembryo
cells until the protodermis was formed. The
presence of this interconnecting network strongly
correlated with tighter adherence of proembryo
cells. The surface network became a common
structure for peripheral proembryo cells being a supramolecular
communication continuum. On the
other hand, embryo-forming cells in the programmed
state when they must be separated
showed limited adherence and distinct contacts via
an external wall structures which were rather selfcovering
(Fig. lf). The presence of interconnecting
fibrillar network provided an early detection of somatic
proembryos within the embryogenic clusters.
Single-cell origin of secondary somatic embryos
from hypocotyl of primary somatic embryos were
demonstrated in repetitive regenerations (Ove~ka
123

M. OVECKA & M. BOBAK
et al. 1997/98). Surface pattern of clustered secondary
somatic proembryos and globular embryos
originating from activated hypocotyl cells (Ove~ka
et al. 1997/98) expressed similar developmentally
regulated cell connection and surface patterning of
proembryos and globular embryos described here.
The protoderm of somatic embryos from the globular
stage with the special cell shape and tight cell adhesion
lacked such external wall structures found
common for all regenerated organs covered by epidermis
(shoots and leaves in organogenic culture).
Our observations of early stages during somatic
embryogenesis are in agreement with spatial and
temporal cell-to-cell communication responsible
for a co-ordinated embryonic pattern expression
before globular stage previously published. Cells in
the embryogenic tissue during development of the
globular embryos were linked with the fibrillar
structure before acquisition of tight adhering cell
contacts, but somatic embryos from the globular
stage were free of such structure in Coffea arabica
(Sondahl et al. 1979), in direct somatic embryogenesis
of Cichorium from root (Dubois et al. 1990,
1992) and leaf tissue (Dubois etal. 1991) and in indirect
somatic embryogenesis of Drosera and Zea
(Samaj et al. 1995). In Cichorium, the extracellular
matrix of tubular structure (Verdus et al. 1993) were
shown of a glycoprotein nature functionally interconnecting
the cytoskeleton (Dubois et al. 1992).
The continuum crossing the cell wall and the
plasma membrane and especially extracellular matrix
within this continuum were precisely characterised
in polarity fixing during Fucus embryogenesis
(see Goodner and Qatrano 1993, Brownlee and
Berger 1995).
Expression and co-ordination of embryonic development
especially establishment and maintenance
of cell division pattern could be (or must be) regulated
by embryo-intrinsic mechanisms requiring effective
signalling machinery in the stage of somatic
proembryo when cells are still undifferentiated.
Cell signalling must be able to block some signals
during cell determination but widely spread other
signals during subsequent development. Stagedependent
regulation ofJ4e epitope expression recognised
by JIM4 antibody was shown in somatic
embryogenesis of carrot (Stacey et aI. 1990). The
epitope expressed in single cultured cells was re-
expressed by cells at the periphery of developing
cell clusters in embryogenesis-promoting conditions
and subsequently the expression in the protoderm
cells of the globular somatic embryo was detected
(Stacey etal. 1990). Developmental and spatial
cell regulation through cell surface epitopes is
documented also by emerging data on involvement
of some AGPs in cell-cell signalling and cellmatrix
interactions during tissue development and
cell proliferation (Duet al. 1996 for a review). In
embryogenic clumps of carrot callus, cell position
and cell type-specific distribution of AGP epitope
at the plasma membrane of the superficial cells of
the clump were shown (Knox 1990). In Cichorium
two embryogenesis-specific proteins absent in
non-embryogenic line were detected and their presence
was correlated with formed external glycoprotein
fibrillar network (Hilbert et al. 1992). From
this point of view any structural components of extracellular
matrix (Sondahl et al. 1979, Goodner
and Qatrano 1993, Brownlee and Berger 1995,
Dubois et al. 1991, 1992, Samaj et al. 1995, and
structures reported here) could be evaluated as an
integral part of the signalling cascade through extracellular
matrix-plasmomembrane-cytoskeleton
continuum (Reuzeau and Point-Lezica 1995) allowing
peripheral proembryo cells to be involved in
co-ordinated development. Temporal occurrence of
external matrix in developmentally restricted stage
indicates its important role in fixing of cell position
and in morphogenesis before the protodermis formation.
The differentiated protodermis was free of
fibrillar and tubular network (Fig. I g, Sondahl et al.
1979, Dubois et al. 1992, Samaj et aL 1995) while
other protoderm-specific marker expressions
within the somatic embryo could be detected such
as EP2 expression in carrot encoding a lipid transfer
protein (Sterk et al. 1991).
Arrangement of embryogenic cell clusters indicate
association of individual proembryos rather than
globular appearance of meristemoids in the organogenic
culture where cell layer specialisation precedes
the shoot regeneration (Ovetka et al. 1997).
Plane of cell division randomly oriented within the
meristemoids was associated with initiated morphogenetic
program of cells activated for regeneration,
still undifferentiated, but keeping a tight cell
adhesion within the meristemoid lacking any inter-
124

CELL SURFACE STRUCTURE AND MORPHOGENESIS IN VITRO
celIular spaces. We have not observed and it has not
been confirmed elsewhere, the presence of external
connecting network on the surface of peripheral
meristemoid cells corresponding to the fibrillar network
around somatic proembryos. The mucilage
surrounding the meristemoid surface was involved
in direct contact with outer walls of peripheral cells
which are regularly thicker compared to internal
cell walls (Fig. 2b, Ove~ka et al. 1997). The mucilage
layer together with thick outer cell wall could
serve as a mechanical separation of meristemoids.
There was clearly demonstrated the correlation between
mucilage accumulation and starch depletion
under unfavourable conditions which indicate mechanical
and metabolic functions of external mucilage
within the meristemoid polysaccharide pool.
In shoot-promoting conditions accumulation of
starch is a prerequisite of the shoot regeneration
(Thorpe and Murashige 1968) and metabolic reserves
are subsequently utilised during shoot primordia
formation without extent mucilage production
(Ove6ka et al. 1997). Plant mucilages are complex
acidic or neutral polysaccharide polymers of
high molecular weight (Fahn 1979) in some cases
having a microfibrillar network texture (Mariani et
al. 1988, Sawidis 1991). Material observed within
the intercellular spaces of parenchyma tissues in redifferentiated
leaves indicating the polysaccharide
nature after PAS reaction (Figs. 4c, d) occurred as a
fibrillar and reticular network mediating intercellular
contacts of the differentiated cells. Communication
within intercellular space containing tissue is
developmentally regulated requiring specialised
modifications of the wall while consequenced by
the loss of cell adhesion (Knox 1992b) and though
probably distinctly regulated from communication
of tightly adherent proembryo cells. Mucilageous
meristemoid structures as well as reticular filling of
intercellular spaces are truly extracellular matrix
(Connolly and Berlyn 1996), however the function
of these materials is probably not comparable with
the function of fibfillar network in early somatic
embryogenesis. On the other hand, co-ordinated
behaviour of the meristemoid cells have been documented.
Meristemoid individuality visible from
early developmental stages to shoot primordia establishment
remaining throughout in the synchrony
of meristemoid cell death indicates the existence of
an effective communication system within the mer-
istemoid although the character and the basic principle
of such communication is still not fully understood.
The morphological description of fine plant cell
wall structure shows a greatimportance of cell adhesion
in early stages of plant regeneration in vitro.
However, more detailed description of fine plant
cell wall structure together with molecular characterisation
of cell surface epitopes is necessary for
the understanding of the role of cell surfaces in the
regulation of cell-to-cell communication and cell
behaviour in asexual plant propagation.
Acknowledgement
This work was partly supported by Grant Academy
VEGA, Grant No. 1/6107/99 from the Slovak
Academy of Sciences.
References
Brownlee C, Berger F. 1995. Extracellular matrix and
pattern in plant embryos: on the look out for developmental
information. Trends in Genetics 11: 344-348.
Connoly JH, Berlyn G. 1996. The plant extracellular
matrix. Canadian Journal of Botany 74: 1545-1546.
Cyr RJ, Bustos MM, Guiltinan M J, Fosket DE. 1987.
Developmental modulation of tubulin protein and
mRNA levels during somatic embryogenesis in cultured
carrot cells. Planta 171: 365-376.
Dijak M, Simmonds DH. 1988. Microtubule organization
during early embryogenesis from mesophyll protoplasts
ofMedicago sativa L. Plant Science 58:183-191.
Domon J-M, Meyer Y, Faye L, David A, David H.
1994. Extracellular (glyco)proteins in embryogenic and
non-embryogenic cell lines of Carribean pine. Comparison
betweem phenotypes of stage one somatic embryos.
Plant Physiology and Biochemistry 32:137-147.
Du It, Clarke AE, Bacie A. 1996. Arabinogalactanproteins:
a class of extracellular matrix proteoglycans involved
in plant growth and development. Trends in Cell
Biology 6:411-414.
Dubois T, Guedira M, Dubois J, Vasseur J. 1990. Direct
somatic embryogenesis in roots of Cichorium: Is
callose an early marker? Annals of Botany 65: 539-545.
Dubois T, Guedira M, Dubois J, Vasseur J. 1991. Direct
somatic embryogenesis in leaves of Cichorium. A
histological and SEM study of early stages. Protoplasma
162: 120-129.
Dubois T, Dubois J, Guedira M, Diop A, Vasseur J.
1992. SEM characterisation of an extracellular matrix
125

M. OVECKA ,~ M. BOBAK
around somatic proembryos in roots of Cichorium. Annals
of Botany 70:119-124.
Dudits D, Brge L, Gyrgyey J. 1991. Molecular and cellular
approaches to the analysis of plant embryo development
from somatic cells in vitro. Journal of Cell Science
99: 475-484.
Emons AMC. 1994. Somatic embryogenesis: cell biological
aspects. Acta Botanica Neerlandica 43: 1-14.
Fahn A. 1979. Secretory tissues in plants. London: Academic
Press.
Goodner B, Quatrano RS. 1993. Fucus embryogenesis:
a model to study the establishment of polarity. Plant
Cell 5: 1471-1481.
Hilbert JL, Dubois T, Vasseur J. 1992. Detection of
embryogenesis-related proteins during somatic embryo
formation in Cichorium. Plant Physiology and Biochemistry
30: 733-741.
Knox JP. 1990. Emerging patterns of organisation at the
plant cell surface. Journal of Cell Science 96: 557-561.
Knox JP. 1992a. Molecular probes for the plant cell surface.
Protoplasma 167: 1-9.
Knox JP. 1992b. Cell adhesion, cell separation and
plant morphogenesis. Plant Journal 2: 137-141.
Mariani P, Raseio N, Baldan B, Paiero P, Urso T.
1988. Epidermal mucilage cells in leaves of Salix species.
Flora 181: 137-145.
Mo LH, Egertsdotter U, yon Arnold S. 1996. Secretion
of specific extracellular proteins by somatic embryos
of Picea abies is dependent on embryo morphology.
Annals of Botany 77:143-152.
Murashige T, Skoog F. 1962. A revised medium for
rapid growth and bioassays with tobacco tissue cultures.
Physiologia Plantarum 15: 473-497.
Ove~ka M, Bobfik M, Erdelsk~ K, gamaj J, Blehovfi
A, Kri]tin J.1996. Morphology and conversion ability
of somatic embryos in long-term embryogenic callus
culture of Papaver somniferum. Biologia (Bratislava)
51: 417-423.
Ove6ka M, Bobfik M, Samaj J. 1997. Development of
shoot primordia in tissue culture of Papaver somniferum
L. Biologia Plantarum 39: 499-506.
Ove~:ka M, Bobfik M, Blehovfi A, Kri~tin J. 1997/98.
Papaver somniferum regeneration by somatic embryogenesis
and shoot organogenesis. Biologia Plantarum
40: 321-328.
Poethig RS. 1989. Genetic mosaics and cell lineage
analysis in plants. Trends Genetics 5: 273-277.
Poulsen GB, Frugis G, Albrechtsen M, Mariotti D.
1996. Synthesis of extracellular proteins in embryogenic
and non-embryogenic cell cultures of alfalfa. Plant Cell
Tissue and Organ Culture 44: 257-260.
Reuzeau C, Pont-Lezica RF. 1995. Comparing plant
and animal extracellular matrix-cytoskeleton connections
- are they alike? Protoplasma 186: 113-121.
Roberts K.1989. The plant extracellular matrix. Current
Opinion in Cell Biology 1: 1020-1027.
Roberts K. 1994. The plant extracellular matrix: in a
new expansive mood. Current Opinion in Cell Biology
6: 688-694.
Sawidis T. 1991. A histochemical study of nectaries of
Hibiscus rosa-sinensis. Journal of Experimental Botany
42: 1477-1487.
Sondahl MR, Salisbury JL, Sharp WR. 1979. SEM
characterization of embryogenic tissue and globular embryos
during high frequency somatic embryogenesis in
coffee callus cells. Zeitschrift fr Pflanzenphysiologie 94:
t85-188.
Stacey N J, Roberts K, Knox JP. 1990. Patterns of expression
of the JIM4 arabinogalactan-protein epitope in
cell cultures and during somatic embryogenesis in Daucus
carota. Planta 180: 285-292.
Sterk P, Booij H, Schellekens GA, Van Kammen A,
De Vries SC. 1991. Cell-specific expression of the carrot
EP2 lipid transfer protein gene. Plant Cell 3: 907-921.
Samaj J, Bobfik M, Erdelsk~ K. 1990. Histologicalanatomical
studies of the structure of the organogenic
callus in Papaver somniferum L. Biologia Plantarum 32:
14-18.
~amaj J, Bobfik M, Blehovfi A, Kri~tin J, Auxtovfi-
Samajovfi O. 1995. Developmental SEM observations
on an extracellular matrix in embryogenic ca]li of Drosera
rotundifolia and Zea mays. Protoplasma I86: 45-49.
Thorpe TA, Murashige T. 1968. Starch accumulation
in shoot-forming tobacco callus cultures. Science 160:
421-422.
Verdus M-C, Dubois T, Dubois J, Vasseur J. 1993.
Ultrastructural changes in leaves of Cichorium during
somatic embryogenesis. Annals of Botany 72: 375-383.
Received May 26, 1998; accepted December 21, 1998
126

Protoplasma (2000) 212:262-267
PROTOPLASMA
9 Springer-Verlag 2000
Printed in Austria
Salt stress induces changes in amounts and localization
of the mitogen-activated protein kinase SIMK in alfalfa roots
Frantigek Balugka 1,2, Miroslav Ovecka 2, and Heribert Hirt 3'*
Institute of Botany, University of Bonn, Bonn, ; Institute of Botany, Slovak Academy of Sciences, Bratislava, and 3 Institute
of Microbiology and Genetics, University of Vienna, Vienna Biocenter, Vienna
Received October 1, 1999
Accepted December 22, 1999
Dedicated to Professor Walter Gustav Url on the occasion of his 70th birthday
Summary. SIMK is an alfalfa mitogen-activated protein kinase
(MAPK) that is activated by salt stress and shows a nuclear localization
in suspension-cultured cells. We investigated the localization
of SIMK in alfalfa (Medicago sati a) roots. Although SIMK was
expressed in most tissues of the root apex, cells of the quiescent
center and statocytes showed much lower SIMK protein amounts.
In cells of the elongation zone, SIMK was present in much higher
amounts in epidermal than in cortex cells. In dividing cells of the
root tip, SIMK revealed a cell cycle phase-dependent localization,
being predominantly nuclear in interphase but associating with the
cell plate and the newly formed cell wall in telophase and early G~
phase. In dividing cells, salt stress resulted in an association of part
of the SIMK with the preprophase band. Generally, salt stress
resulted in much higher amounts of SIMK in dividing cells of
the root apex and epidermal cells of the elongation zone. These
data demonstrate that amounts and subcellular localization of
SIMK in roots is highly regulated and sensitive to environmental
stress.
Keywords: Mitogen-activated protein kinase; Salt stress; Osmotic
stress; Medicago sati a.
Introduction
Hyperosmotic-stress signaling in yeasts, mammals, and
plants is mediated through highly conserved MAPK
cascades (Brewster et al. 1993, Galcheva-Gargova
et al. 1994, Han et al. 1994, Munnik et al. 1999). MAP
kinase pathways are found in all eukaryotes, including
* Correspondence and reprints: Institute of Microbiology and
Genetics, University of Vienna, Vienna Biocenter, Dr.-Bohr-Gasse
9, A-1030 Vienna, Austria.
E-mail: hehi@gem.univie.ac.at
plants, and are involved in transducing a variety of
extracellular signals including growth factors, hormones,
as well as biotic and abiotic stresses (Jonak
et al. 1999, Waskiewicz and Cooper 1995). MAPK cascades
are usually composed of three protein kinases
that upon activation undergo sequential phosphorylation
(Robinson and Cobb 1997). By phosphorylation
of conserved threonine and tyrosine residues, a MAPK
becomes activated by a specific MAPK kinase
(MAPKK). A MAPKK kinase (MAPKKK) activates
MAPKK through phosphorylation of conserved
threonine and/or serine residues. MAPK pathways
may integrate a variety of upstream signals through
interaction with other kinases or G proteins (Robinson
and Cobb 1997). The latter factors often directly
serve as coupling agent between a plasma-membranelocated
receptor protein that senses an extracellular
stimulus and a cytoplasmic MAPK module. At the
downstream end of the module, activation of the cytoplasmic
MAPK module often induces the translocation
of the MAPK into the nucleus, where the kinase
activates certain sets of genes through phosphorylation
of specific transcription factors (Treisman 1996).
In other cases, a given MAPK may translocate to
other sites in the cytoplasm to phosphorylate specific
enzymes (protein kinases, phosphatases, lipases, etc.)
or cytoskeletal components (Cohen 1997, Robinson
and Cobb 1997). By tight regulation of MAPK localization
and through expression of certain signaling
components and substrates in particular cells, tissues,
or organs, particular MAPK pathways can mediate

E Balus"ka et al.: Mitogen-activated protein kinase SIMK in alfalfa 263
signaling of a multitude of extracellular stimuli and
bring about a large variety of specific responses.
We have investigated whether hyperosmotic stress
in plants is also mediated by MAP kinase pathways.
We have found that SIMK (stress-inducible MAP
kinase) (originally named MsK7; Jonak et al. 1993) is
activated by hyperosmotic stress in alfalfa cells
(Munnik et al. 1999). In suspension-cultured cells,
SIMK was found to be a constitutively nuclear protein
and was not undergoing changes in its intracellular
location upon salt stress. In this report, we have investigated
the presence and localization of SIMK in
alfalfa roots. Our results indicate that the tissuespecific
pattern and subcellular localization of SIMK
is influenced by both cell cycle cues and environmental
conditions.
Material and methods
Plant material and sample preparation
Seeds of Medicago sati a L. cv. Europa were placed on moist filter
paper in petri dishes and germinated for 3 days in cultivating chambers
in darkness at 25 ~ For salt stress treatment roots of 3-dayold
seedlings were incubated in 0 or 200 mM NaC1. 8 mm long root
tips were fixed in 3.7% formaldehyde in stabilizing buffer [SB;
50 mM piperazine-N, N9-bis(2-ethanesulfonic acid), 5 mM MgSO4,
5 mM EGTA, pH 6.9] for 1 h. After rinsing in SB and phosphatebuffered
saline (PBS), root tips were dehydrated in a graded ethanol
series diluted with PBS. The root tips were then infiltrated with
Steedman's wax diluted in absolute ethanol in step with the proportion
1 : 1 (ethanol and wax, v/v), followed by 2 steps in pure
wax. The infiltrated root tips were allowed to polymerize in pure
wax at room temperature overnight. All chemicals, if not stated
otherwise, were obtained from Sigma Chemical Co. (St. Louis, Mo.,
U.S.A.).
Antibody production and specificity
M23 antibody was produced against a synthetic peptide, encoding
t]~e carboxyl-terminal amino acids FNPEYQQ of SIMK (Jonak
et al. 1993). The antibody was found to monospecifically detect
SIMK, and preincubation of the antibody with an excess of the
synthetic petide completely blocked immunofluorescent labeling of
SIMK in suspension-cultured cells (Munnik et al. 1999) and root
sections (data not shown).
Indirect immunofluorescence microscopy
7 mm thick median longitudinal sections were prepared from the
fixed root tips and mounted on slides coated with glycerol-albumen
(Serva, Heidelberg, Federal Republic of Germany). Sections
expanded in a drop of distilled water, and they were allowed to
adhere to the slides. For easy penetration of antibodies, the sections
were first dewaxed in ethanol, rehydrated in an ethanol and PBS
series, and kept in PBS for 30 min After a 10 rain methanol treatment
at -20 ~ and a SB rinse for 30 min at room temperature, the
sections were incubated with affinity-purified SIMK-specific antibody
diluted 1 : 1000 with PBS for 60 rain at room temperature.
After a rinse in SB the sections were incubated with secondary fluorescein
isothiocyanate-conjugated anti-rabbit immunoglobulin G
raised in goat (Sigma) diluted 1 : 100 with PBS for 60 min at room
temperature in darkness. Nuclear DNA was stained with 49,6-
diamidino-2-pheuylindole (DAPI) diluted at 1 g/ml PBS for 10 rain.
After rinsing with PBS, the sections were stained with 0.01%
Toluidine Blue in PBS for 10 rain and mounted in antifade mountant
medium containing p-phenylenediamine. Immunofluorescence
observation was examined with an Axiovert 405M inverted light
microscope (Zeiss, Oberkochen, Federal Republic of Germany) and
Olympus BX 50 (Olympus, Tokyo, Japan) microscope equipped with
epifluorescence and standard FITC exciter and barrier filters (BP
450-490 nm wavelength, LP 520 nm wavelength). Photographs were
taken on Kodak T-Max films rated at 400 ASA.
Results
Cell cycle phase-dependent localization of SIMK
SIMK encodes a MAPK that was recently identified
to be activated by salt stress in suspension-cultured
cells, leaves, and roots (Munnik et al. 1999). In suspension-cultured
cells, salt stress was not found to
alter the subcellular localization of SIMK. However,
suspension-cultured cells are constantly exposed to
mechanical stress (shaking), conditions that were
found to activate SIMK, albeit to a much smaller
extent than salt stress (Munnik et al. 1999). To study
the behavior and localization of SIMK in a more
natural system, roots of 3-day-old alfalfa seedlings
were analyzed before and after salt stress. Figure 1
shows immunofluorescence pictures taken from longitudinal
sections of nonstressed root tips. To correlate
SIMK localization with the cell cycle stages, the sections
were also stained with DAPI and compared. By
this analysis, proliferating cells of the meristematic
region revealed a predominantly nuclear localization
of SIMK in interphase cells. In accordance with
the cell cycle phase-dependent expression found in
suspension-cultured cells (Jonak et al. 1993), G2 phase
cells (Fig. 1 A) had higher amounts of SIMK protein
than cells in other stages. Interestingly, in mitosis and
early-Gl-phase cells, S1MK was found to accumulate
in the newly formed nuclei as well as with the assembling
cell plates and young cell walls (Fig. 1 A). These
results prompted us to perform a more thorough
analysis of the localization of SIMK during different
mitotic stages. In prophase, metaphase, and anaphase,
SIMK was clearly excluded from the condensing chromosomes
and the spindle apparatus (Fig. 1 C-J). In
telophase (lower cell in Fig. 1 K, L), SIMK started to
accumulate at the periphery of the reforming nuclei
(Fig. 1 K). At a slightly later stage of mitosis (upper cell

264 F. Balugka et al.: Mitogen-activated protein kinase SIMK in alfalfa
Fig. 1A-P. Cell cycle phase- and salt-stress-dependent localization of SIMK. A-L Root tip ceils of the proliferation zone that represent
different cell cycle stages. A and B SIMK is a nuclear protein in nonstressed interphase ceils. Stars in A indicate cells in the G2 phase of
the cell cycle. C and D During prophase, SIMK has disappeared from the nucleus and is now excluded from both condensing chromatids
and the spindle apparatus (arrowheads in C). E and F Early metaphase (indicated by star). G and H Metaphase (indicated by star); SIMK
is still excluded from chromatids and the spindle apparatus. I and l Anaphase (indicated by star). K and L Telophase (lower cell in K and
L); SIMK is located at the periphery of newly forming nuclei (arrows in K). The phragmoplast is shown by an asterisk in K. After cytokinesis
(upper cell in K and L), SIMK has re-entered the daugtber nuclei (stars in K) but is also associated with the new cell wall (arrowhead
in K). M-O Cells of the proliferation zone from salt-stressed roots (200 mM NaC1, 1 h). M and N Salt-stressed interphase cells show
SIMK as a nuclear protein, but preprophase bands (arrows in M) are also labeled. O and P In some salt-stressed telophase cells, SIMK is
localized to phragmoplasts but not to the cell plate (asterisk in O). A, C, E, G, I, K, M, and O are fluorescence micrographs of sections that
were decorated with SIMK-specific antibody. B, D, F, H, J, L, N, and P are the corresponding sections that were stained with DAPI. Bar in
P: for A-N, 5 gm; for O and R 8 am

E Balus'ka et al.: Mitogen-activated protein kinase SIMK in alfalfa 265
in Fig. 1 K, L), SIMK was seen to have re-entered the
nuclei (Fig. 1 K), but a clear association with the young
cell wall was also apparent (Fig. 1 K). The association
of SIMK with young cell walls was consistently found
to extend well into the G1 phase, when cells had clearly
finished cytokinesis, as can be seen by inspecting the
cells in Fig. 1 A, E, and I.
Salt-stress-associated changes in intracellular
localization of SIMK
A concentration of 200 mM NaC1 induces the activation
of SIMK in suspension-cultured cells, leaves, and
roots (Munnik et al. 1999). Root tip cells of alfalfa
seedlings that were treated at this salt concentration
for t h showed little changes in the intracellular localization
of SIMK. In interphase cells (Fig. 1M, N),
SIMK was found to be still predominantly nuclear.
However, some cells also showed an association of
SIMK with preprophase bands (Fig. 1M). Whereas
prophase and metaphase cells showed a SIMK staining
pattern that was similar to nonstressed cells (data
not shown), some salt-treated root tip cells in telophase
(Fig. 1 O, P) revealed colocalization of SIMK
with phragmoplasts (Fig. 1 O). In these cells, SIMK
labeling of the newly forming cell walls was distinctly
absent.
Tissue-specific expression of SIMK influenced
by salt stress
When root tips were analyzed for SIMK expression at
a lower magnification, it became apparent that SIMK
was abundantly present in most tissues of the root
tip (Fig. 2 A). An exception to this rule were the quiescent
center and the statocytes at the very root apex
(Fig. 2A) showing much lower amounts of SIMK.
Interestingly, root tips of salt-stressed seedlings
showed much higher SIMK protein amounts (Fig.
2B). In the transition zone of unstressed roots, SIMK
showed considerably higher expression levels in the
epidermis than in the adjacent cortex (Fig. 2C), and
salt stress exacerbated this difference even further
(Fig. 2 D). When cells of the elongation region were
analyzed after salt stress, a large percentage of epidermal
cells was observed to have undergone plasmolysis
(Fig. 2E). In these cells, the protoplast
retained most of SIMK protein, but some SIMK
protein was also found to be associated with the
plasma membrane and the cell wall (Fig. 2 E).
Discussion
In animals, yeasts, and plants, specific MAP kinase
pathways are involved in mediating responses to
hyperosmotic stress. The family of mammalian MAP
kinases including the SAPK(stress-activated protein
kinase)/JNK(Jun N-terminal kinase)/p38 is activated
by hyperosmotic as well as various other types of stress
(Waskiewicz and Cooper 1995). In Saccharomyces
cerevisiae, the HOG1 MAP kinase pathway is exclusively
used for mediating hyperosmotic stress (Brewster
et al. 1993). In alfalfa plants, the SIMK pathway is
involved in signaling hyperosmotic stress (Munnik
et al. 1999).
Activation of MAP kinases often involves the
nuclear import of the MAP kinase from the cytoplasm
after activation. Studies on the mammalian ERK1/
ERK2 kinases showed that the phosphorylation of the
MAP kinase by the upstream MAP kinase kinase was
an essential step to induce the nuclear translocation
(Chen et al. 1992, Lenormand et al. 1993). Recent data
revealed that MAP kinases can shuttle between cytoplasmic
and nuclear compartments, and that the time
spent in any one compartment can be influenced by a
number of parameters that may differ in a systemspecific
way. In mammalian cells, MAPKKs may act as
cytoplasmic anchors for inactive MAPKs (Fukuda
et al. 1997). Phosphorylation of MAPK is thought to
induce dissociation from the MAPKK, thereby allowing
nuclear import of the MAPK. Other evidence
suggests that phosphorylation-induced dimerization
may also contribute to nuclear import of MAPKs
(Khokhlatchev et al. 1998). Investigations of the Schizosaccharomyces
pornbe Spcl/Styl stress-signaling
MAP kinase pathway revealed that the nuclear target
of the Spcl/Styl kinase, the transcription factor Atfl,
plays an active role in retaining the MAPK in the
nucleus (Gaits et al. 1998). Finally, studies on the
HOG1 kinase in budding yeast added more complexity
by showing that besides regulation of entry and
anchoring of the MAPK in the nucleus, nuclear export
of the MAPK also contributes to the overall time of
MAPK nuclear residence (Ferrigno et al. 1998).
To study the intracellular localization of SIMK, indirect
immunofluorescence microscopy was performed
on root tip sections of seedlings that were either
untreated or stressed with 200 mM NaC1 for different
times. Our analysis in nonstressed roots revealed a cell
cycle phase-dependent localization of SIMK, showing
constitutive nuclear staining during interphase. The

266 E Balu~ka et al.: Mitogen-activated protein kinase SIMK in alfalfa
Fig. 2A-E. Salt-stress-induced changes in tissue-specific expression of SIMK in root apices. A With the exception of the quiescent center
and statocytes, SIMK is abundantly present in all other tissues in nonstressed root apices. B Salt stress results in enhanced amounts of
SIMK in all tissues of the root apex. C In the transition zone of nonstressed alfalfa roots, SIMK shows preferential epidermal expression.
D Salt stress (200 mM NaCI, 1 h) results in enhanced levels of SIMK protein in the epidermis. To visualize the SIMK staining pattern in
epidermal cells, micrograph D was exposed for a much shorter time than micrograph C. SIMK labeling of cortex cells in D was similar to
that seen in C. E Salt-stress-induced plasmolyzed cells of the elongation zone. SIMK is still mostly present in the nuclei but also shows
clear labeling of material that is associated with the plasma membrane and cell walls. A-E are fluorescence micrographs of sections that
were decorated with SIMK-specific antibody. Bar in E: for A and B, 20 gm; for C-E, 7 gm
early mitotic stages revealed SIMK to be excluded
from the mitotic spindle and chromatids, whereas
during telophase and early G1, the most obvious localization
was found with the assembling cell plate and
with the new cell wall. In fact, we observed close colocalization
of SIMK with myosin VIII in alfalfa root
apices (data not shown). Myosin VIII was shown to
accumulate at assembling cell plates and young cell
walls of cytokinetic root cells of diverse plant species
(Reichelt et al. 1999). Since myosin activity was shown
to be regulated via a MAP kinase pathway in animal
cells (Klemke et al. 1997), it could be interesting to test
if plant myosin VIII is a target of the SIMK cascade.
SIMK was also found to give more intense labeling
in cells in the G2 phase, supporting our previous results
that SIMK steady-state mRNA levels fluctuate with
the cell cycle, giving maximum levels in the G2 phase
(Jonak et al. 1993). A possible role of SIMK during the
cell cycle is also supported by the finding that cells of
the quiescent center contained much lower amounts
of SIMK protein than cells that were actively proliferating.
However, more direct evidence is required to
support the idea that SIMK has a function in the cell
cycle.
Salt stress was also found to affect the localization
of SIMK. Although interphase cells retained SIMK in

E Balugka et al.: Mitogen-activated protein kinase SIMK in alfalfa 267
the nucleus, some late-G2 cells showed association of
SIMK with preprophase bands. This could be brought
about by translocation of some of the nuclear SIMK
protein. Alternatively, because cell fractionation of
suspension-cultured cells had shown that SIMK is
present in both the nucleus and the cytoplasm
(Munnik et al. 1999), SIMK association with preprophase
bands may occur by recruitment of the
kinase from a diffuse cytoplasmic pool. Whatever the
mechanism, most investigations have so far revealed
that MAPKs are extremely dynamic molecules, and
green-fluorescent-protein fusion technology may be
necessary to follow changes in MAPK localization in
real time. Besides affecting the subcellular localization
of SIMK in dividing cells, the major effect of salt stress
was found in increased SIMK protein levels in both
dividing cells of the root apex and nondividing cells of
the transition and elongation zone. It remains to be
analyzed whether the increased amounts of SIMK in
these cells after salt stress are due tO increased gene
expression, mRNA, or protein stability.
Overall, our data show that both expression as well
as localization of SIMK in root tips is regulated in a
highly complex manner. Besides revealing a cell cycle
phase dependency, subcellular localization of SIMK
was also sensitive to salt stress. In addition to sal t
stress, SIMK may also be involved in sensing of other
environmental factors like temperature and mechanical
stress (unpubl. data). Future experimental work
will be necessary to reveal the physiological significance
of environmental-stress-induced changes in
amounts and intracellular localization of SIMK.
Although our present knowledge is scarce, the SIMK
pathway appears to be well suited to study the intracellular
dynamics of stress signaling in plants.
Acknowledgments
The work was supported grants from the Austrian Science Foundation
(P12188-GEN and P11729-GEN) and from the TMR program
of the European Union. This work was also supported by Grant
Agency VEGA, grant nr. 3009 from the Slovak Academy of Sciences.
Financial support to AGRAVIS by the Deutsche Agentur fiir
Raumfahrtangelegenheiten (Bonn, Federal Republic of Germany)
and the Ministerium fiir Wissenschaft und Forschung (D~sseldorf,
Federal Republic of Germany) is gratefully acknowledged.
References
Chen RH, Sarnecki C, Blenis J (1992) Nuclear localisation and
regulation of ERK- and RSK-endcoded protein kinases. Mol Cell
Biol 12:915-927
Cheng M, Boulton TG, Cobb MH (1996) ERK3 is a constitutively
nuclear protein kinase. J Biol Chem 271:8951-8958
Cohen P (1997) The search for physiological substrates of MAP and
SAP kinases in mammalian cells. Trends Cell Biol 7:353-361
Ferrigno P, Posas E Koepp D, Saito H, Silver PA (1998) Regulated
nucleo/cytoplasmic exchange of HOG1 MAPK requires the
importin [3 homologs NMD5 and XPO1. EMBO J 17:5606-5614
Fukuda M, Gotoh Y, Nishida E (1997) Interaction of MAP kinase
with MAP kinase kinase, its possible role in the control of nucleocytoplasmic
transport of MAP kinase. EMBO J 16:1901-1908
Gaits F, Degols G, Shiozaki K, Russell P (1998) Phospohorylation
and association with the transcription factor Atfl regulate localisation
of Spcl/Styl stress-activated kinase in fission yeast. Genes
Dev 12:1464-1473
Galcheva-Gargova Z, Derijard B, Wu I, Davis R (1994) An
osmosensing signal transduction pathway in mammalian cells.
Science 265:806-808
Han J, Lee J-D, Bibbs L, Ulevitch RJ (1994) A MAP kinase targeted
by endotoxin and hyperosmolarity in mammalian cells. Science
265:808-811
Jonak C, Ligterink W, Hirt H (1999) MAP kinases in plant signal
transduction. Cell Mol Life Sci 55:204-213
- Pgty A, B6gre L, Hirt H, Heberle-Bors E (1993) The plant
homolog of MAP kinase is expressed in a cell cycle-dependent
and organ specific manner. Plant J 3:611-617
Khokhlatchev AV, Canagarajah B, Wilsbacher J, Robinson M, Atkinson
M, Goldsmith E, Cobb MH (1998) Pbosphorylation of the
MAP kinase ERK2 promotes its homodimerization and nuclear
localisation. Cell 93:605-615
Klemke RL, Cai S, Giannini AL, Gallagher PJ, de Lanerolle P,
Cheresh DA (1997) Regulation of cell motility by mitogenactivated
protein kinase. J Cell Biol 137:481-492
Lenormand R Sardet C, Pages G, L'Allemain G, Brunet A, Pouyssegur
J (1993) Growth factors induce nuclear translocation of MAP
kinases (p42mapk and p44mapk) but not of their activator MAP
kinase kinase (p45mapkk) in fibroblasts. J Cell Biol 122: 1079-
1088
Marshall CJ (1995) Specificity of receptor tyrosine kinase signaling:
transient versus sustained extracellular signal-regulated kinase
activation. Cell 80:179-185
Munnik T, Ligterink W, Meskiene I, Calderini O, Beyerly J, Musgrave
A, Hirt H (1999) Distinct osmosensing protein kinase pathways
are involved in signaling moderate and severe hyperosmotic
stress. Plant J 20:381-388
Murashige T, Skoog F (1962) A revised medium for rapid growth
and bioassays with tobacco tissue culture. Physiol Plant 15:
473-479
Reichelt S, Knight AE, Hodge TR Balu~ka F, Samaj J, Volkmann D,
Kendrick-Jones J (1999) Characterization of the unconventional
myosin VIII in plant cells and its localization at the postcytokinetic
cell wall. Plant J 19:555-569
Robinson MJ, Cobb MH (1997) Mitogen-activated protein kinase
pathways. Curr Opin Cell Biol 9:180-186
Treisman R (1996) Regulation of transcription by MAP kinase cascades.
Curr Opin Cell Biol 8:205-215
Waskiewicz AJ, Cooper J (1995) Mitogen and stress response pathways,
MAP kinase cascades and phosphatase regulation in
mammals and yeast. Curr Opin Cell Biol 7:798-802
Brewster JL, deValoir T, Dwyer ND, Winter E, Gustin MC (1993)
An osmosensing signal transduction pathway in yeast. Science
259:1760-1763

The EMBO Journal Vol. 21 No. 13 pp. 3296±3306, 2002
Involvement of the mitogen-activated protein kinase
SIMK in regulation of root hair tip growth
Jozef SÏ amaj 1 ,2 , Miroslav Ovecka 1 ,3,4 ,
Andrej Hlavacka 3 , Fatma Lecourieux 1 ,
Irute Meskiene 1 , Irene Lichtscheidl 5 ,
Peter Lenart 1 ,JaÂn Salaj 2 , Dieter Volkmann 3 ,
La szlo BoÈ gre 6 , FrantisÏek BalusÏka 3 ,4 and
Heribert Hirt 1 ,7
1 Institute of Microbiology and Genetics, Vienna Biocenter, University
of Vienna, Dr Bohrgasse 9, A-1030 Vienna, 5 Institut of Ecology,
University of Vienna, Althanstrasse 14, A-1091 Vienna, Austria,
3 Institute of Botany, Plant Cell Biology Department, University of
Bonn, Kirschallee 1, D-53115 Bonn, Germany, 2 Institute of Plant
Genetics and Biotechnology, Slovak Academy of Sciences,
Akademicka 2, PO Box 39A, SK-950 07 Nitra, 4 Institute of Botany,
Slovak Academy of Sciences, DuÂbravska cesta 14, SK-842 23
Bratislava, Slovak Republic and 6 School of Biological Sciences,
Royal Holloway, University of London, Egham TW20 0EX, UK
7 Corresponding author
e-mail: hehi@univie.ac.at
Mitogen-activated protein kinases (MAPKs) are
involved in stress signaling to the actin cytoskeleton in
yeast and animals. We have analyzed the function of
the stress-activated alfalfa MAP kinase SIMK in root
hairs. In epidermal cells, SIMK is predominantly
nuclear. During root hair formation, SIMK was activated
and redistributed from the nucleus into growing
tips of root hairs possessing dense F-actin meshworks.
Actin depolymerization by latrunculin B resulted in
SIMK relocation to the nucleus. Conversely, upon
actin stabilization with jasplakinolide, SIMK co-localized
with thick actin cables in the cytoplasm.
Importantly, latrunculin B and jasplakinolide were
both found to activate SIMK in a root-derived cell
culture. Loss of tip-focused SIMK and actin was
induced by the MAPK kinase inhibitor UO 126 and
resulted in aberrant root hairs. UO 126 inhibited targeted
vesicle traf®cking and polarized growth of root
hairs. In contrast, overexpression of gain-of-function
SIMK induced rapid tip growth of root hairs and
could bypass growth inhibition by UO 126. These data
indicate that SIMK plays a crucial role in root hair tip
growth.
Keywords: actin cytoskeleton/MAP kinase/root hairs/
signaling/tip growth
Introduction
Mitogen-activated protein kinases (MAPKs), a speci®c
class of serine/threonine protein kinases, are involved in
controlling many cellular functions in all eukaryotes. A
general feature of MAPK cascades is their composition of
three functionally linked protein kinases. An MAPK is
phosphorylated and thereby activated by a MAPK kinase
(MAPKK), which itself becomes activated by another
serine/threonine protein kinase, a MAPKK kinase
(MAPKKK). Targets of MAPKs can be various transcription
factors, protein kinases or cytoskeletal proteins
(Whitmarsh and Davis, 1998).
Signaling through MAPK cascades is involved in cell
differentiation, division and stress responses (Robinson
and Cobb, 1997). In plants, a number of studies have
demonstrated that MAPKs are signaling biotic and abiotic
stresses, including cold and drought (Jonak et al., 1996),
wounding (Seo et al., 1995; BoÈgre et al., 1997; Zhang and
Klessig, 1998) and plant±pathogen interactions (Ligterink
et al., 1997; Cardinale et al., 2000; NuÈhse et al., 2000).
Moreover, plant MAPKs are also involved in cell division
and hormone action (Ligterink and Hirt, 2001).
Studies in animal and yeast cells revealed distinct roles
of various MAPKs such as dynamic organization of the
actin cytoskeleton and polarization of cell growth
(Mazzoni et al., 1993; Zarzov et al., 1996; Rousseau
et al., 1997; SchaÈfer et al., 1998; Delley and Hall, 1999).
Prominent examples of MAPKs controlling the actin
cytoskeleton are the p38 kinase of animal cells and MPK1
of budding yeast. Both p38 and MPK1 are critical for
polarization of the actin cytoskeleton (Mazzoni et al.,
1993; Zarzov et al., 1996; Rousseau et al., 1997). The p38
kinase is also responsible for cell migration of animal cells
(Rousseau et al., 1997) and MPK1 is involved in yeast cell
growth (Mazzoni et al., 1993; Zarzov et al., 1996).
Recently, the stress-induced MAPK (SIMK) (Munnik
et al., 1999) and its upstream activator SIMKK (Kiegerl
et al., 2000) have been characterized and shown to be
inducible by osmotic stress and various fungal elicitors
(Cardinale et al., 2000). Here, we studied the function of
SIMK during root hair formation. In trichoblasts, SIMK
was located to peripheral spots predicting root hair
outgrowth. In growing root hairs, SIMK was found at
root hair tips together with F-actin meshworks. After
treatment of root hairs with actin drugs and the MAPKK
inhibitor UO 126 (Favata et al., 1998), changes in the actin
cytoskeleton were correlated with changes in the subcellular
localization and activity of SIMK. Moreover,
overexpression of gain-of-function SIMK in transgenic
plants resulted in increased root hair formation and
growth. Our data suggest that SIMK plays an important
role in root hair tip growth linking polar growth to MAPK
signaling and the actin cytoskeleton.
Results
Tip-focused SIMK localization in growing
root hairs
In situ hybridization with a SIMK antisense probe revealed
that SIMK was strongly expressed in alfalfa root hairs
(data not shown). The polyclonal M23 antibody was
derived against the heptapeptide FNPEYQQ, correspond-
3296 ã European Molecular Biology Organization

Stress MAP kinase SIMK in plant tip growth
Fig. 1. Immunoblot and immuno¯uorescence detection of total and active SIMK. (A) Root extracts were prepared and immunoblotted with actin antibody
(lane 1) or with SIMK antibody M23 (lane 2). (B) Salt treatment of roots for 10 min activated SIMK as revealed by immunoblotting crude root
extracts with phospho-speci®c SIMK antibody N103 (lane 2) and SIMK-speci®c antibody M23 (lane 3). Active SIMK is hardly detected in control
roots with N103 (lane 1). (C) Immuno¯uorescence microscopy of SIMK in elongating root cells of M.sativa L. using the Steedman's wax embedding
technique. Note that SIMK is localized predominantly to nuclei (indicated by arrowheads), but depleted from nucleoli (indicated by stars). (D) DIC
image of (C). (E) Immunodepletion control of epidermal root cells (shown in F) with M23 after pre-incubation with FNPEYQQ heptapeptide.
(F) Corresponding DIC image for (E). (G) Trichoblast before root hair initiation showing cell periphery-associated spot-like SIMK labeling at the
outer tangential cell wall (arrows). (H) Trichoblast at the bulging stage: SIMK labeling appears at the outermost domain of the developing bulge
(arrows). (I) Growing root hair showing SIMK labeling focused to the tip (arrows) and in spot-like structures along the root hair tube. SIMK is depleted
from the nucleus and nucleoli (arrowhead and star, respectively). (J) Root epidermal cells showing very low levels of active SIMK labeled with
N103 antibody. (K) Corresponding DIC image for (J). Nuclei and nucleoli in (J) and (K) are indicated by arrowheads and stars. (L) Tip of a growing
root hair showing accumulation of active SIMK in spot-like structures at the root hair tip (arrows). (M) Immunodepletion control of root hair with
N103 after pre-incubation with CTDFMTpEYpVVTRWC peptide. Bar = 15 mm for (C±F), 10 mm for (G±K) and 5 mm for (L) and (M).
ing to the C-terminus of SIMK (Cardinale et al., 2000),
and speci®cally recognizes SIMK but not other related
MAPKs (Munnik et al., 1999; Cardinale et al., 2000).
Immunoblot analysis of root extracts revealed that M23
recognized a single band of 46 kDa that corresponds to
SIMK (Figure 1A, lane 2). A monoclonal actin antibody
used in this study reacts speci®cally with a single band of
45 kDa in crude root cell extracts (Figure 2A, lane 1). A
phospho-speci®c polyclonal antibody N103 was raised in
rabbit against CTDFMTpEYpVVTRWC peptide of
3297

J.SÏ amaj et al.
Fig. 2. Co-immunolocalization of tubulin and SIMK (A±C) or actin and SIMK (D±O) in root hairs using the freeze-shattering technique.
(A) Microtubules are organized in longitudinal and net-axially arranged arrays in non-growing parts of the root hair tube and are much less abundant
in subapical and apical zones of growing root hair apices. (B) SIMK accumulates in root hair apices and in distinct spots. (C) Merged image indicating
no signi®cant co-localization (yellow color) of microtubules and SIMK at root hair tips and within root hair tubes. Arrows indicate root hair tip.
(D±F) Control growing root hairs. (G±I) Growing root hairs treated with 10 mM latrunculin B (LB) for 30 min. (J±L) Growing root hairs treated with
5 mM jasplakinolide (JK) for 60 min. (M±O) Growing root hairs treated with 50 mM brefeldin A (BFA) for 60 min. (D) Dense actin meshworks are
present at root tips, and F-actin organizes in the form of longitudinal bundles further away from the root hair tip. (E) SIMK accumulation in root hair
apices and in distinct spots further away from the hair tip. (F) Co-localization (yellow color) of actin and SIMK at root hair tips (indicated by arrowheads).
Nuclei are indicated by arrows in (E) and (F). (G) LB disrupts F-actin in growing root hairs and depletes actin from root hair tips (arrowheads).
(H) SIMK relocates from tips (arrowheads) to nuclei (arrows) upon LB treatment. (I) DAPI staining of (H). Nuclei are indicated by arrows. (J) JK
induces F-actin stabilization and the appearance of thick actin cables protruding to root hair tips. (K) SIMK is located to thick cables and to roundshaped
cytoplasmic spots after JK treatment. (L) Extensive co-localization (yellow) of SIMK with thick F-actin cables in JK-treated hairs. (M) BFA
causes the disappearance of the F-actin meshwork from the tip (arrowhead) while F-actin ®laments deeper in the cytoplasm remain intact. (N) SIMK
is relocated from the tip and concentrates in patches within the cytoplasm in BFA-treated hairs. (O) SIMK patches are associated with actin ®laments
(arrows) in BFA-treated hairs. Bar = 25 mm for (D±F), 30 mm for (G±I) and 15 mm for (A), (B) and (J±O).
SIMK. The N103 antibody was puri®ed on protein A and
immunoaf®nity columns. Because SIMK is activated by
salt stress (Munnik et al., 1999), protein extracts prepared
from salt-treated roots were immunoblotted with N103
antibody. In untreated roots, very little active SIMK was
detected by N103 (Figure 1B, lane 1). Upon salt stress,
N103 speci®cally recognized a 46 kDa band (Figure 1B,
lane 2) corresponding to SIMK as detected by the speci®c
SIMK antibody M23 (Figure 1B, lane 3). In protoplasts cotransformed
with SIMK and its activator SIMKK (Kiegerl
3298

Stress MAP kinase SIMK in plant tip growth
et al., 2000), N103 speci®cally recognized activated SIMK
(data not shown). These data show that N103 antibody is
suitable for studying activated SIMK.
In root tips, SIMK is predominantly nuclear in stele and
cortex tissues, but is more abundant in epidermal cells
(BalusÏka et al., 2000b). Cytological analysis of longitudinal
sections from the root transition and elongation
zones revealed that SIMK was mostly present in nuclei of
epidermal cells (arrows in Figure 1C and D). SIMK
showed a spot-like nuclear staining, but was not detected
in nucleoli (stars in Figure 1C). An immunodepletion
control in which the SIMK antibody was pre-incubated
with the FNPEYQQ heptapeptide con®rmed the speci®city
of labeling in roots (Figure 1E and F).
When root hair formation was analyzed, SIMK was
found to accumulate in outgrowing bulges and at root hair
tips (Figures 1G±I and 2E). SIMK accumulated in distinct
spots at the cell periphery facing the outer tangential cell
wall of trichoblast (Figure 1G) marking the site of bulge
outgrowth. During bulge formation, SIMK showed strong
association with cell periphery-associated spots of the
bulge (Figure 1H). In growing root hairs, SIMK was found
to accumulate within root hair tips and in spot-like
structures in the root hair tube (Figure 1I). This pattern
of SIMK distribution along root hairs was con®rmed by
semi-quantitative measurements of ¯uorescence intensity
(data not shown). Using phospho-speci®c N103 antibody,
activation of SIMK was observed during root hair
formation. In root epidermal cells, only very weak labeling
was found (Figure 1J and K). In growing root hairs, active
SIMK accumulated in root hair tips in the form of distinct
spots (Figure 1L). An immunodepletion control in which
the N103 antibody was pre-incubated with the
CTDFMTpEYpVVTRWC peptide revealed no labeling
in roots and con®rmed the speci®city of N103 antibody
(Figure 1M). These data show that accumulation of active
SIMK in root hair tips correlates with root hair formation.
SIMK co-localizes with F-actin meshworks in
root hair tips and with F-actin cables after
jasplakinolide treatment
Actin ®laments, but not microtubules, are abundant at tips
of growing root hairs (Bibikova et al., 1999; Braun et al.,
1999; Miller et al., 1999) and are involved in root hair
initiation and polar growth (BalusÏka et al., 2000a).
Because SIMK can associate with mitotic but not cortical
microtubules of dividing cells under certain circumstances
(BalusÏka et al., 2000b), we also investigated whether
SIMK co-localized with microtubules in growing root
hairs. We did not ®nd co-localization of SIMK with
microtubules in root hair apices and in root hair tubes
(Figure 2A±C). In order to study a possible SIMK
association with actin ®laments, we performed co-localization
studies of actin and SIMK in growing root hairs
(Figure 2D±F) and root hairs treated with actin drugs
(Figure 2G±L). SIMK co-localized with the dense F-actin
meshwork at root hair tips (Figure 2F, arrowheads) and in
spots along longitudinal cables of actin ®laments in
subapical and deeper portions of growing root hairs.
These data suggest that SIMK is associated with both the
actin cytoskeleton and vesicular compartments in root
hairs.
To gain further insight into the importance of F-actin
and apical SIMK localization, roots were treated with
latrunculin B (LB), an inhibitor of actin polymerization
(Gibbon et al., 1999; BalusÏka et al., 2000a), or with
jasplakinolide, an F-actin-stabilizing drug (Bubb et al.,
2000). We performed time course experiments with these
drugs followed by co-localization of actin and SIMK
within bulges and root hairs. After treating roots for 30 min
with LB, effective depolymerization of F-actin was
observed in bulging trichoblasts. Under these conditions,
SIMK disappeared from periphery-associated spots of root
hair bulges and relocated to the nucleus (data not shown).
In growing root hairs treated with LB, ¯uorescent spots
presumably representing G-actin or patches of fragmented
F-actin were distributed evenly over the entire length of
the root hairs (Figure 2G). In root hair apices, LB
treatment resulted in disintegration and depletion of
dense apical F-actin meshworks (Figure 2G). The most
intense SIMK labeling of LB-treated root hairs was found
invariably within nuclei and nucleoli (Figure 2H),
as revealed by nuclear 4¢,6-diamidino-2-phenylindole
(DAPI) staining (Figure 2I).
To study the association of SIMK with F-actin in more
detail, root hairs were treated with jasplakinolide.
Jasplakinolide-induced stabilization of F-actin was accompanied
by the appearance of thick F-actin cables protruding
into the extreme root hair tips (Figure 2J). In
jasplakinolide-treated root hairs, SIMK co-localized
extensively with these thick F-actin cables (Figure 2K
and L). Besides this, SIMK was also located to individual
spots within the cytoplasm (Figure 2K and L). These data
suggest that drugs affecting the organization and dynamics
of the F-actin cytoskeleton (the G-actin/F-actin ratio) have
a direct impact on the intracellular localization of SIMK.
SIMK is associated with vesicular traf®c in
root hairs
Besides co-localization with tip-focused F-actin meshworks,
SIMK was found in distinct spots along the entire
length of the root hair (Figures 1 and 2). To investigate
further the vesicle-associated localization of SIMK, we
used brefeldin A (BFA) as an ef®cient inhibitor of
vesicular traf®cking in plant cells (Satiat-Jeunemaitre
et al., 1996). Upon BFA treatment, the apical dense
meshworks of F-actin in growing root hairs disappeared
(Figure 2M) and SIMK was located throughout the root
hair cytoplasm in spotty and patchy structures of variable
sizes (Figure 2N). These BFA-induced patches were
distributed along F-actin (Figure 2O), indicating that the
actin cytoskeleton might be important for their formation.
The tip-focused gradient of SIMK and F-actin was lost in
root hairs treated with BFA (Figure 2N and O). These
results suggest that vesicular traf®cking is involved in the
control of SIMK distribution in root hair tips.
Jasplakinolide and latrunculin B induce activation
of SIMK in cultured root cells
In order to investigate a possible role for the actin
cytoskeleton on SIMK activity, we tested two actin drugs
for their effects on SIMK activity in root-derived suspension
cultured cells. Immunokinase analysis revealed that
both jasplakinolide (5 mM) and LB (10 mM) activate SIMK
(Figure 3, uper panel). After treating cells with LB, SIMK
3299

J.SÏ amaj et al.
Fig. 3. Immunokinase analysis of the jasplakinolide (JK)- and the latrunculin
B (LB)-induced activation of SIMK in root-derived suspension-cultured
cells. Alfalfa cells were treated with 5 mM JK or 10 mM
LB for the indicated times. Extracts from treated cells, containing
100 mg of total protein, were immunoprecipitated with 5 mg of
protein A-puri®ed SIMK antibody. Kinase reactions were performed
with 1 mg/ml MBP as a substrate, 0.1 mM ATP and 2 mCi of
[g- 32 P]ATP. Phosphorylation of MBP was analyzed by autoradiography
after SDS±PAGE. Corresponding controls with DMSO [DMSO was
used at the same concentrations as for dilution of jasplakinolide
(0.25%) and latrunculin B (0.1%), respectively] showing no kinase
activity at 0, 10, 30 and 60 min are presented in the lower panels. An
immunoblot showing constant levels of SIMK protein during treatments
with actin drugs is presented in the lowest panel.
was activated within 10 min, and activity decreased at later
time points. In contrast, jasplakinolide treatment resulted
in activation of SIMK after 10 min, but maximal activation
was observed at 60 min. Control treatment of cells with 0.1
and 0.25% dimethylsulfoxide (DMSO), the concentrations
used for applying the actin drugs, showed no SIMK
activation. Importantly, SIMK protein levels were constant
during treatment with both LB and jasplakinolide
(Figure 3, lower panel). These data show that the state of
the actin cytoskeleton affects the activity of SIMK.
Overexpression of active SIMK affects root hair
formation and growth
In order to see whether SIMK is necessary for root hair
growth, both gain-of-function and loss-of-function constructs
of SIMK (SIMK-GOF and SIMK-LOF) were
produced and stably expressed in transgenic tobacco. In
analogy to the Drosophila rolled MAPK-GOF mutant
(Brunner et al., 1994), the D348N exchange was introduced
within the a-L16 helix in the C-terminal part of the
protein. This region is involved in binding to upstream
kinases and phosphatases. Transient expression assays in
protoplasts showed that SIMK-GOF protein was more
active than wild-type SIMK (data not shown). Biochemical
analysis on stably transformed tobacco plants by M23
immunokinase assays and immunoblotting revealed that
four SIMK-GOF lines showed clearly increased MAPK
activity in comparison with control plants, while protein
levels were similar in both control and SIMK-GOF plants
(Figure 4A, upper panel). Since M23 equally recognizes
SIMK and its tobacco homolog SIPK/Ntf4 (Wilson et al.,
1998), M23 immunokinase assays determine both endogenous
SIPK/Ntf4 and ectopically expressed SIMK-GOF
in tobacco plants.
In a similar approach, SIMK-LOF transgenic tobacco
lines were produced. The K69M point mutation within the
b-3 sheet of subdomain II creates an inactive loss-offunction
MAPK enzyme by disrupting the ATP-binding
site (Robinson et al., 1996). Both in transient protoplast
assays (data not shown) and in transgenic SIMK-LOF
tobacco lines (Figure 4A, lower panel), no signi®cant
changes in MAPK activity and protein levels were
observed.
Inspection of transgenic tobacco SIMK-GOF plants
revealed that the root hair formation zone was extremely
shortened and root hairs were much longer (Figure 4B)
in comparison with control non-transformed roots
(Figure 4C). In contrast, roots of SIMK-LOF mutant
plants showed no visible root hair phenotype (Figure 4D).
Taken together, these results indicate that overexpression
of active SIMK stimulates root hair growth and formation.
MAPKK inhibitor UO 126 inhibits root hair growth
Since overexpression of inactive SIMK-LOF was ineffective
in inhibiting root hair formation and growth, we
used a MAPKK inhibitor to down-regulate MAPK activity
in root hairs. For this purpose, we treated roots with UO
126, an inhibitor of MAPK activation (Favata et al., 1998).
After a 1 h treatment, the ®rst morphological changes in
emerging and growing root hairs became evident and
correlated with inhibition of root hair tip growth. After 2 h
of UO 126 treatment, ballooning of emerging root hairs
(Figure 5A) and swelling of apical and basal parts of
growing root hairs (Figure 5D) were observed. No such
effect was found using UO 124, an inactive analog of UO
126 (Figure 5B and E), or DMSO at the same concentration
as a solvent for UO 126 and UO 124, respectively
(Figure 5C and F). These data suggest that MAPK activity
is necessary for tip growth of root hairs.
UO 126 affects actin organization and SIMK
distribution within root hairs
To test whether MAPK activity is directly involved in the
regulation of actin organization, roots were treated with
UO 126. After UO 126 treatment, vacuolation of root hair
tips (Figure 6A and B) was observed. At the same time, we
noticed the partial destruction of F-actin cables located
deeper within hairs and actin re-arrangement around
newly formed vacuoles (Figure 6A). Under these conditions,
SIMK was distributed evenly in cytoplasm and
nuclei of treated hairs without any preferable accumulation
in subcellular compartments (Figure 6B). In contrast,
treatment with the non-active analog UO 124 caused no
signi®cant changes in tip-focused actin and SIMK localization
(Figure 6C±E). These experiments indicate that
inhibition of MAPK activity causes remodeling of the
actin cytoskeleton in emerging root hairs.
UO 126 inhibits root hair growth by changing
polar vesicular traf®cking and root hair
cytoarchitecture
To study the effect of UO 126-induced tip growth
inhibition in root hairs in real time, we used videoenhanced
microscopy. In control alfalfa root hairs, the
apex (5±7 mm) was usually ®lled with small vesicles which
were densely packed (Figure 7A) and showed random
motions with a radius of 0.5±1 mm. Occasionally, several
vesicles moved in a row over longer distances (3±8 mm) in
the axial direction and also in various other directions
3300

Stress MAP kinase SIMK in plant tip growth
Fig. 4. Effect of overexpression of SIMK-GOF and SIMK-LOF on kinase activity and root hair formation. (A) MAPK activity and protein levels in
control (c) tobacco SR1 plants and transformed SIMK-GOF (1±4) and SIMK-LOF (1 and 2) SR1 lines using M23 recognizing both SIMK and its
tobacco homolog SIPK/Ntf4. (B) Root of a SIMK-GOF tobacco plant showing a very short root hair formation zone (indicated by a bracket) and
much longer root hairs (indicated by arrow) as compared with (C and D). (C and D) Roots of non-transformed and SIMK-LOF plants, respectively,
showing normal root hair formation zones (brackets) and regular length of root hairs (arrows). Bar = 400 mm for (B±D).
within the apical dome. Whereas large organelles were
slow, vesicles moved with velocities between 4 and
6 mm/s. At the plasma membrane, secretory vesicles are
supposed to fuse and deliver their contents to the cell wall,
whereas endocytotic vesicles should appear at the ¯anks of
the apical dome. These processes cannot be observed,
however, due to the small size and fast movement of the
vesicles.
The ®rst signs of an effect of inhibition of MAPK
activity appeared after 6±8 min of treatment with 10 mM
UO 126 (Figure 7B). At this time, tip growth slowed down
(from 4±6 to 1.5±2 mm/min) and ®nally stopped within
15 min. At 6±8 min, the movement of the vesicles became
hectic and disorganized in the tip region and the amplitude
of random displacements increased (1±2 mm), whereas
targeted movements over longer distances disappeared
(Figure 7C). Importantly, putative endocytotic vesicles at
the plasma membrane became visible in cells treated with
UO 126 because they stayed at apical dome ¯anks for 6±8 s
before they disappeared (Figure 7C, H and L). In the tube,
vesicles slowed down and, after 15 min, became immobile
and static. Tubular small vacuoles within the tip
(Figure 7D) in¯ated to large roundish vacuoles that
oscillated in random motion (Figure 7E). Their tonoplasts
were lined by vesicles and these vacuoles often fused
(Figure 7E and F). During this process, the shape of the
apical dome became swollen over the whole length where
apical secretory vesicles showed disorganized random
motions and the cell wall appeared thicker (Figure 7D±F).
Similar effects of UO126 on root hair cytoarchitecture and
vesicular traf®cking were detected in root hairs of control
tobacco plants where UO 126 caused growth arrest within
5 min (Figure 7G±I). These results suggest that UO 126
inhibits root hair tip growth by modifying cytoarchitecture
and polar vesicular targeting and traf®cking. Importantly,
the root hairs of tobacco SIMK-GOF transgenic plants
continued to grow in the presence of UO 126 (growth
was inhibited by 10±15% only) and maintained root
hair cytoarchitecture and normal vesicular traf®cking
(Figure 7J and K). In comparison with control plants,
growth of SIMK-GOF root hairs was inhibited only at a
10-fold higher concentration of UO 126 (Figure 7L),
showing minor changes in cytoarchitecture and no large
and round vacuoles appearing in their apices (Figure 7L).
These results show that overexpression of active SIMK
can override inhibition of root hair growth by UO 126.
3301

J.SÏ amaj et al.
Fig. 5. Morphological changes on emerging (A) and growing root hairs (D) induced by 2 h treatment with 10 mM UO 126, a MAPKK inhibitor.
(A) UO 126 causes swelling of emerging root hairs (arrows). (B and C) Emerging root hairs are not affected by 2 h treatment with 10 mM UO 124 or
0.1% DMSO, respectively. (D) UO 126 induces vacuolation and swelling of apices of growing root hairs (arrows). (E and F) UO 124 (10 mM) and
(0.1%) DMSO have no morphological effect on growing root hairs. Bar = 30 mm for (A), (D) and (F), and 40 mm for (B), (C) and (E).
Discussion
The stress-activated MAP kinase SIMK is strongly
expressed in root hairs. The selective enrichment of
active SIMK in tips of emerging root hairs coincides
with dynamic F-actin meshworks (Braun et al., 1999;
BalusÏka et al., 2000a; BalusÏka and Volkmann, 2002).
Depolymerization and stabilization of F-actin activates
SIMK, indicating that MAPK activity is directly affected
by altering F-actin dynamics. Inhibition of MAPKK
activity caused changes in the subcellular distribution of
actin and SIMK, resulting in tip growth inhibition and
aberrant root hair morphology. Conversely, overexpression
of active SIMK resulted in enhanced growth of root
hairs. These data indicate that SIMK functions in root hair
growth.
Active SIMK is present in root hairs
In situ hybridization and immunolocalization revealed that
both SIMK transcript and protein are present within root
hairs. Previously, we have shown that SIMK is localized
predominantly to nuclei in meristematic cells of root
apices (BalusÏka et al., 2000b). SIMK is also found in
nuclei of elongating epidermal root cells (Figure 1).
During bulge formation and root hair formation, SIMK is
relocated polarly from nuclei towards bulging domains of
trichoblasts and tips of growing root hairs, SIMK is also
located at root hair tips in an active form.
MPK1 activation in yeast occurs by a weakening of the
cell wall associated with stretch-stressed plasma membranes
(Kamada et al., 1995), with concominant actin repolarization
(Delley and Hall, 1999). A similar mechanism
can also be envisaged for plant tip growth within
Fig. 6. Immunolocalization of actin (green, A and C) and SIMK (red,
B and D) in root hairs treated with 10 mM UO 126, a MAPKK inhibitor
(A and B), or with 10 mM UO 124, an inactive analog of UO 126
(C±E) for 60 min. (A) Note the vacuolation at the tips (arrows) and
F-actin depletion from root hairs. The remaining actin labeling is
evenly distributed. (B) SIMK is distributed evenly in root hair cytoplasm
and nuclei, except holes represented by vacuoles (arrows).
(C) The tip actin meshwork (arrows) and ®lamentous actin are preserved
in hairs treated with UO 124. (D) SIMK is tip focused in hairs
treated with UO 124 (arrows). (E) Co-localization of SIMK with actin
meshworks at the tip (arrows, yellow color). Thick F-actin bundles
within trichoblasts are indicated by arrowheads. Bar = 30 mm for (A)
and (B), 18 mm for (C) and (D) and 23 mm for (E±G).
outgrowing bulges and root hair tips, both representing
weak cell periphery domains (BalusÏka et al., 2000a). In
order to relieve the high stretch stress imposed on the
plasma membrane, both abundant exocytosis (Kell and
Glaser, 1993; Fricker et al., 2000) and dense F-actin
meshworks (Ko and McCulloch, 2000) are essential,
3302

Stress MAP kinase SIMK in plant tip growth
Fig. 7. Video-enhanced microscopy of growing root hairs treated with UO 126. (A) An untreated growing alfalfa root hair with a vesicle-rich apical
dome. (B±F) Alfalfa root hair treated with 10 mM UO 126. Note that UO 126 markedly changes the cytoarchitecture and shape of the root hair apex
causing inhibition of root hair tip growth. (G±I) Tobacco growing root hair treated with 10 mM UO 126. This concentration of UO 126 caused growth
arrest within 5 min and similar changes in the cytoarchitecture and shape of the root hair apex to those in alfalfa (A±F). (J±L) Tobacco SIMK-GOF
root hair treated with 10 mM UO 126 (J and K) or 100 mM UO 126 (L) for the indicated time points. Note that 10 mM UO126 did not cause growth
arrest and changes in root hair cytoarchitecture and shape within 60 min, while 100 mM UO 126 inhibits root hair growth within 60 min, causing
minor changes in the cytoarchitecture. Bar = 7 mm for (A±F) and 5 mm for (G±L).
culminating in the onset of root hair tip growth (Braun
et al., 1999; Miller et al., 1999; BalusÏka et al., 2000a).
SIMK localization and activity are associated with
actin organization
Dense F-actin meshworks at root hair tips of different plant
species were observed in root hairs by immunolabeling
with actin antibodies or in vivo using green ¯uorescent
protein (GFP) fused to the F-actin-binding domain of talin
(Braun et al., 1999; BalusÏka et al., 2000a; BalusÏka and
Volkmann, 2002). An intact actin cytoskeleton is necessary
for root hair tip growth because root hairs treated with
the F-actin disruptors LB or cytochalasin D are inhibited in
their growth and show morphological abnormalities
(Miller et al., 1999; BalusÏka et al., 2000a; Ovecka et al.,
2000). F-actin recruitment towards root hair bulges and
growing tips roughly coincides with local accumulations
of actin-depolymerizing factor (ADF) (Jiang et al., 1997),
pro®lin (Braun et al., 1999; BalusÏka et al., 2000a) and Rop
GTPases (Molendijk et al., 2001). SIMK was found to
localize within tips of growing root hairs. Tip-focused
localization of SIMK disappeared after treatment with LB
and resulted in nuclear accumulation of SIMK. The
association of SIMK with ®lamentous actin could be
enhanced by jasplakinolide, an inducer of actin polymerization.
These ®ndings indicate that an intact actin
cytoskeleton is necessary for the proper localization of
SIMK within root hair tips. Interestingly, both actin drugs
activate SIMK in suspension-cultured root cells, suggesting
that the state and dynamics of the actin cytoskeleton
directly in¯uence SIMK activity. Actin reorganization is
likely to be mediated by actin-binding proteins (e.g.
pro®lin, ADF, villin, ARP, ®mbrin and calponin), and
SIMK might phosphorylate and regulate one of these
actin-binding proteins. Both ADF and pro®lin are actinbinding
proteins responsible for the dynamics of F-actin
meshworks (for a review on plant cells see Staiger, 2000).
Presently, it is not clear which actin cytoskeletal proteins
are targeted by SIMK in vivo. In animal cells, p38
regulates the dynamic organization of the actin cytoskeleton
via phosphorylation of the small heat shock protein
HSP27 (SchaÈfer et al., 1998). HSP27 acts as an F-actincapping
protein inhibiting polymerization of G-actin in its
phosphorylated state (Benndorf et al., 1994). Interestingly,
3303

J.SÏ amaj et al.
p38 is homologous to the yeast HOG1 MAP kinase that colocalizes
with actin patches in osmotically stressed yeast
cells (Ferrigno et al., 1998). Recently, it was shown that
extracellular regulated kinase interacts with the actin and
calponin homology domains of actin-binding proteins
(Leinweber et al., 1999).
Our results show that SIMK is activated in response to
F-actin destabilization, suggesting that SIMK might have a
direct role in transmitting signals from the actin cytoskeleton
to the tip growth machinery. Disturbances of actin
dynamics, induced with either LB or jasplakinolide,
activated SIMK. Interestingly, Gachet et al. (2001)
reported that LB activates stress-activated MAP kinase
(SAPK) of ®ssion yeast. This F-actin-dependent SAPK
cascade is part of a new mitotic checkpoint ensuring
proper spindle orientation. Similarly, activation of the
yeast MAP kinase MPK1 via LB-mediated depolymerization
of F-actin triggers the morphogenesis checkpoint in
budding yeast cells (Harrison et al., 2001). One possibility
is that depolymerization of F-actin imposes mechanical
stress on the cellular architecture which activates MAPK
cascades in order to regain the balance of intracellular
forces (Chicurel et al., 1998). This is supported by our
in vivo observation when inhibition of MAPK activity
leads to changes in cytoarchitecture and actin-dependent
vesicular motilities within root hair tips. However,
®lamentous actin can also inhibit the activity of certain
kinases, as has been shown recently for the c-Abl tyrosine
kinase in mammalian cells (Woodring et al., 2001). This is
probably not the case for SIMK in plant cells because
stabilization of F-actin with jasplakinolide activates
SIMK. Possibly, SIMK monitors the balance of forces,
and activation of SIMK could be necessary for the
recovery of a balanced cytoarchitecture (Chicurel et al.,
1998). It would be interesting to test this idea in yeast
where MAPKs monitor actin-dependent mitosis and
morphogenesis checkpoints (Gachet et al., 2001;
Harrison et al., 2001).
UO 126 inhibits root hair growth and disrupts
polar actin and SIMK distribution
UO 126 inhibited root hair growth and affected the
subcellular organization of both the actin cytoskeleton and
SIMK in root hairs. As revealed by video-enhanced
microscopy, the disturbance of actin-dependent polar
vesicle traf®cking by UO 126 is most likely responsible
for the swelling and diffuse growth of root hairs.
Gain-of-function SIMK induces longer root hairs
The way in which the D334N mutation acts in the
Drosophila MAPK rolled mutant is not well understood,
but it results in a gain-of-function phenotype in the
signaling pathway of Drosophila eye development
(Brunner et al., 1994). Exchanging the homologous
amino acid in SIMK, we observed higher MAPK activity
in transformed protoplasts and plants. A possible explanation
comes from analyzing a similar mutation in a yeast
FUS3 allele where it was concluded that FUS3 was more
active because the MAPK phosphatases were less able to
inactivate the FUS3 mutant kinase (Hall et al., 1996).
Overexpression of SIMK-GOF in tobacco plants showed a
phenotype of long root hairs correlated with sustained
activity of SIMK. On the other hand, overexpression of
SIMK-LOF in tobacco showed no visible root hair
phenotype. This result could be explained either by
functional redundancy of several MAPKs in root hairs
or, alternatively, by the inability of SIMK-LOF to compete
suf®ciently with the endogenous tobacco MAPK homolog
(SIPK/Ntf4) to inhibit root hair growth. In support of
functional redundancy, knockout mutants of AtMPK6, the
Arabidopsis homolog of SIMK, show no phenotype
(K.Shinozaki, personal communication).
SIMK and vesicular traf®c in root hairs
Root hairs are tubular extensions of trichoblasts growing
exclusively at their tips by means of highly polarized exoand
endocytosis (Shaw et al., 2000). Exocytotic vesicles
deliver cell wall polysaccharides such us pectins and
xyloglucan (Sherrier and VandenBosch, 1994), and molecules
such as lectins and arabinogalactan proteins were
detected in cell walls of root hair tips (Diaz et al., 1989;
S Ï amaj et al., 1999). On the other hand, nothing is known
about endocytosis in root hairs which should be responsible
for plasma membrane and cell wall recycling. By
immunolocalization, active SIMK was found to accumulate
in spots within root hair tips which are known to be
®lled with exo- and endocytotic vesicles (Sherrier and
VandenBosch, 1994; this study). We show here that the
MAPKK inhibitor UO 126 inhibits targeted vesicular
traf®c, resulting in growth arrest of root hairs. Since
overexpression of SIMK-GOF in tobacco could overcome
growth arrest by UO 126, it is likely that SIMK has an
important function in growth control of root hairs. Because
of resolution limits of video-enhanced microscopy and the
dynamic behavior of abundant exocytotic vesicles, it is not
possible to conclude that all exocytotic vesicles fuse to the
tip plasma membrane. The vesicles could also undergo a
`kiss and run' mode of movement in delivering cell wall
material to the tip. Diffuse growth and cell wall thickening
at tips of UO 126-inhibited root hairs suggest that
exocytosis is not blocked. On the other hand, after growth
arrest with UO 126, the putative endocytotic vesicles
appearing on tip ¯anks persist at the plasma membrane for
a much longer time. These observations suggest that
MAPK activity may be required for endocytosis rather
than for exocytosis in growing root hairs, but clari®cation
of this issue awaits further advances in plant endocytosis
research, and future work will have to determine the exact
molecular mechanism of the interdependent regulation of
actin and SIMK and their roles in polar vesicular
traf®cking.
Materials and methods
Plant material
Seeds of Medicago sativa L. cv. Europa were placed on moist ®lter paper
in Petri dishes and germinated for 3 days in cultivating chambers in
darkness at 25°C. Three-day seedlings with straight primary roots
40±50 mm long were selected.
For in vivo observation, M.sativa and Nicotiana tabacum L. cv. SR1
(control SIMK-LOF and SIMK-GOF) seeds were sterilized and
germinated on half-strength MS medium or on wet ®lter paper for 2 or
3 days. Seedlings were mounted between slide and coverslip that
contained three layers of para®lm as spacers. Within such chambers, the
primary roots could develop for at least 24 h. This method ensured that the
roots could adapt to the liquid growth medium, and avoided mechanical
stress during mounting of roots on the microscope. Chambers were placed
upright into a Petri dish ®lled with Fahraeus growth medium (Fahraeus,
3304

Stress MAP kinase SIMK in plant tip growth
1957). After placing the chambers under the microscope, root hairs of
short to medium length were selected and their growth was measured for
at least 10 min. When elongation was normal, i.e. in the range of
0.4±0.6 mm/min, the growth medium was removed with ®lter paper and
replaced by Fahraeus growth medium containing UO 126.
Suspension cultures were prepared from callus derived from alfalfa
roots (M.sativa L.) and maintained as described (Baier et al., 1999). Log
phase cells were used 3 days after a weekly 1:3 dilution in fresh medium.
Vector constructs
The SIMK-GOF mutation was created by changing D348 into N348, and
the SIMK-LOF mutation was created by changing K84 into M84. Sitedirected
mutagenesis was performed in pALTER vector (Clontech). For
recloning to plant expression vectors, PCR was performed with the
following primers: SIMK 5¢-NcoI 5¢-AAAACCATGGAAGGAGGA-
GGAGC-3¢ and SIMK 3¢-NotI 5¢-AAAGGATCCTCAGCGGCC-
GCCCTGCTGGTACTCAGGGTTAAATGC-3¢.
PCR products were recloned to pBIN-HygTX vector (provided by
C.Gatz, IGF Berlin) with the TMV omega leader as SmaI±XbaI inserts.
Plant transformation
Leaf discs of the SR1 tobacco line were transformed with Agrobacterium
containing binary vector with SIMK-GOF or SIMK-LOF. Nine and four
primary transformants, respectively, were regenerated from SR1 tobacco
transformed with SIMK-GOF and SIMK-LOF.
Treatments with inhibitors
For dilution of drugs, distilled water or modi®ed Fahraeus medium were
used. Root apices were treated with 100 mM BFA (Sigma). LB (10 mM;
Calbiochem) in 0.1% DMSO and 5 mM jasplakinolide (Molecular Probes)
in 0.25% DMSO were used as actin inhibitors. For UO 126 and UO 124
inhibitors (Calbiochem), concentrations ranging from 10 nM up to
100 mM were used.
Antibodies and immunoblotting
Antibodies against SIMK (M23, recognizing the FNPEYQQ heptapeptide
of SIMK) (Cardinale et al., 2000), phosphorylated SIMK (N103,
recognizing the double-phosphorylated CTDFMTpEYpVVTRWC peptide
of SIMK) and actin (clone C4, Amersham) were tested on crude root
extracts as described by BoÈgre et al. (1997). Immunoreactive bands were
visualized using enhanced chemiluminescence (ECL kit, Amersham)
according to the manufacturer's instructions. Membranes were exposed to
Biomax X-ray ®lm (Kodak) for 30 s.
In vitro kinase activity assays
Cell extracts containing 100 mg of total protein were immunoprecipitated
overnight with 5 mg of protein A-puri®ed SIMK antibody. The
immunoprecipitated kinase was washed three times with wash buffer
(50 mM Tris±HCl pH 7.4, 250 mM NaCl, 5 mM EGTA, 0.1% Tween-20,
5 mM NaF) and once with kinase buffer (20 mM HEPES pH 7.4, 15 mM
MgCl 2 , 5 mM EGTA, 1 mM dithiothreitol). Kinase reactions were
performed as described (BoÈgre et al., 1997). Brie¯y, the immunocomplexes
were incubated for 30 min at room temperature in 15 ml of kinase
buffer containing 1 mg/ml MBP, 0.1 mM ATP and 2 mCi of [g- 32 P]ATP.
The reaction was stopped by adding SDS±PAGE loading buffer, and the
phosphorylation of MBP was analyzed by autoradiography after
SDS±PAGE.
Immunolocalization and microscopy
Immunolocalizations were performed using both freeze-shattering (Braun
et al., 1999) and Steedman's wax embeding methods (BalusÏka et al.,
2000a) except that half-strength stabilization buffer was used for sample
®xation. Immuno¯uorescence images were collected using a Leica
confocal microscope TCS4D (Leica, Heidelberg, Germany) or an
Axioplan 2 microscope (Zeiss, Oberkochen, Germany).
Video microscopy of living root hairs
Video microscopy was performed as described by Foissner et al. (1996).
Brie¯y, the bright®eld image from an Univar microscope (Reichert-
Leica, Austria) equipped with a 403 planapo objective was collected
with a high-resolution video camera (Chalnicon C 1000.1, Hamamatsu,
Germany), processed by a digital image processor (DVS 3000,
Hamamatsu Germany) and recorded on digital video tape (JVC).
Images and video scenes were imported into a personal computer
(Sony Vaio).
Acknowledgements
We thank Markus Braun for his advice and help with freeze shattering.
This work was supported by an EU Marie Curie individual fellowship to
J.SÏ. and the EU Sokrates Program to A.H. Financial support to F.B. and
D.V. by the Deutsches Zentrum fuÈr Luft- und Raumfahrt (Bonn,
Germany) is highly appreciated. J.S Ï ., J.S. and F.B. receive partial support
from the Slovak Academy of Sciences, Grant Agency Vega (grants no. 2/
6016/99 and 2031), Bratislava, Slovakia. We thank the Alexander von
Humboldt foundation (Bonn, Germany) for donations of equipment. The
work was also supported by grants to H.H. from the Austrian Science
Foundation.
References
Baier,R., Schiene,K., Kohring,B., Flaschel,E. and Niehaus,K. (1999)
Alfalfa and tobacco cells react differently to chitin oligosaccharides
and Sinorhizobium meliloti nodulation factors. Planta, 210, 157±164.
BalusÏka,F. and Volkmann,D. (2002) Actin-driven polar growth of plant
cells. Trends Cell Biol., 12, 14.
BalusÏka,F., Salaj,J., Mathur,J., Braun,M., Jasper,F., SÏamaj,J., Chua,
N.-H., Barlow,P.W. and Volkmann,D. (2000a) Root hair formation:
F-actin-dependent tip growth is initiated by local assembly of pro®linsupported
F-actin meshworks accumulated within expansin-enriched
bulges. Dev. Biol., 227, 618±632.
BalusÏka,F., Ovecka,M. and Hirt,H. (2000b) Salt stress- and cell cycle
phase-dependent changes in expression and subcellular localisation of
the alfalfa mitogen-activated protein kinase SIMK. Protoplasma, 212,
262±267.
Benndorf,R., Hayess,K., Ryazantsev,S., Wieske,M., Behlke,J. and
Lutsch,G. (1994) Phosphorylation and supramolecular organization
of murine small heat shock protein HSP25 abolish its actin
polymerization-inhibiting activity. J. Biol. Chem., 269, 20780±20784.
Bibikova,T.N., Blanca¯or,E.B. and Gilroy,S. (1999) Microtubules
regulate tip growth and orientation in root hairs of Arabidopsis
thaliana. Plant J., 17, 657±665.
BoÈgre,L., Ligterink,W., Meskiene,I., Barker,P.J., Heberle-Bors,E.,
Huskisson,N.S. and Hirt,H. (1997) Wounding induces the rapid and
transient activation of a speci®c MAP kinase pathway. Plant Cell, 9,
75±83.
Braun,M., BalusÏka,F., von Witsch,M. and Menzel,D. (1999)
Redistribution of actin, pro®lin and phosphatidylinositol-4,5-
bisphosphate (PIP 2 ) in growing and maturing root hairs. Planta,
209, 435±443.
Brunner,D., Oellers,N., Szabad,J., Biggs,W.H., Zipursky,L.S. and
Hafen,E. (1994) A gain of function mutation in Drosophila MAP
kinase activates multiple receptor tyrosine kinase signaling pathways.
Cell, 76, 875±888.
Bubb,M.R., Spector,I., Beyer,B.B. and Fosen,K.M. (2000) Effects of
jasplakinolide on the kinetics of actin polymerization. An explanation
for certain in vivo observations. J. Biol. Chem., 275, 5163±5170.
Cardinale,F., Jonak,C., Ligterink,W., Niehaus,K., Boller,T. and Hirt,H.
(2000) Differential activation of four speci®c MAPK pathways by
distinct elicitors. J. Biol. Chem., 275, 36734±36740.
Chicurel,M.E., Chen,C.S. and Ingber,D.E. (1998) Cellular control lies in
the balance of forces. Curr. Opin. Cell Biol., 10, 232±239.
Delley,P.-A. and Hall,M.N. (1999) Cell wall stress depolarizes cell
growth via hyperactivation of RHO1. J. Cell Biol., 147, 163±174.
Diaz,C., Melchers,L., Hooykaas,P., Lugtenberg,B. and Kijne,J. (1989)
Root lectin as a determinant of host-plant speci®city in the
Rhizobium±legume symbiosis. Nature, 338, 579±581.
Fahraeus,G. (1957) The infection of clover root hairs by nodule bacteria
studied by a simple glass slide technique. J. Gen. Microbiol., 16,
374±381.
Favata,M.F. et al. (1998) Identi®cation of a novel inhibitor of mitogenactivated
protein kinase kinase. J. Biol. Chem., 273, 18623±18632.
Ferrigno,P., Posas,F., Koepp,D., Saito,H. and Silver,P.A. (1998)
Regulated nucleo/cytoplasmic exchange of HOG1 MAPK requires
the importin b homologs NMD5 and XPO1. EMBO J., 17, 5606±5614.
Foissner,I., Lichtscheidl,I.K. and Wasteneys,G.O. (1996) Actin-based
vesicle dynamics and exocytosis during wound wall formation in
Characean internodal cells. Cell Motil. Cytoskel., 35, 35±48.
Fricker,W., Jarvis,M.C. and Brett,C.T. (2000) Turgor pressure,
membrane tension and the control of exocytosis in higher plants.
Plant Cell Environ., 23, 999±1003.
3305

J.SÏ amaj et al.
Gachet,Y., Tournier,S., Millar,J.B.A. and Hyams,J.S. (2001) A MAP
kinase-dependent actin checkpoint ensures proper spindle orientation
in ®ssion yeast. Nature, 412, 352±355.
Gibbon,B.C., Kovar,D.R. and Staiger,C.J. (1999) Latrunculin B has
different effects on pollen germination and tube growth. Plant Cell,
11, 2349±2363.
Hall,P.J., Cherkasova,V., Elion,E., Gustin,M.C. and Winter,E. (1996)
The osmoregulatory pathway represses mating pathway activity in
Saccharomyces cerevisiae: isolation of a FUS3 mutant that is
insensitive to the repression mechanism. Mol. Cell. Biol., 16,
6715±6723.
Harrison,J.C., Bardes,E.S.G., Ohya,Y. and Lew,D.J. (2001) A role for
the Pkc1p/Mpk1p kinase cascade in the morphogenesis checkpoint.
Nat. Cell Biol., 3, 1±5.
Jiang,C.-J., Weeds,A.G. and Hussey,P.J. (1997) The maize actindepolymerizing
factor, ZmADF3, redistributes to the growing tip of
elongating root hairs and can be induced to translocate into the
nucleus with actin. Plant J., 12, 1035±1043.
Jonak,C., Kiegerl,S., Ligterink,W., Barker,P.J., Huskisson,N.S. and
Hirt,H. (1996) Stress signaling in plants: a mitogen-activated protein
kinase pathway is activated by cold and drought. Proc. Natl Acad. Sci.
USA, 93, 11274±11279.
Kamada,Y., Jung,U.S., Piotrowski,J. and Levin,D.E. (1995) The protein
kinase C-activated MAP kinase pathway of Saccharomyces cerevisiae
mediates a novel aspect of the heath shock response. Genes Dev., 9,
1559±1571.
Kell,A. and Glaser,R.W. (1993) On the mechanical and dynamic
properties of plant cell membranes: their role in growth, direct gene
transfer and protoplast fusion. J. Theor. Biol., 160, 41±62.
Kiegerl,S. et al. (2000) SIMKK, a mitogen-activated protein kinase
(MAPK) kinase, is a speci®c activator of the salt stress-induced
MAPK, SIMK. Plant Cell, 12, 2247±2258.
Ko,K.S. and McCulloch,C.A.G. (2000) Partners in protection:
interdependence of cytoskeleton and plasma membrane in
adaptations to applied forces. J. Membr. Biol., 174, 85±95.
Leinweber,B.D., Leavis,P.C., Grabarek,Z., Wang,C.L.A. and
Morgan,K.G. (1999) Extracellular regulated kinase (ERK)
interaction with actin and the calponin homology (CH) domain of
actin-binding proteins. Biochem. J., 344, 117±123.
Ligterink,W. and Hirt,H. (2001) Mitogen-activated protein (MAP)
kinase pathways in plants: versatile signaling tools. Int. Rev. Cytol.,
201, 209±215.
Ligterink,W., Kroj,T., zur Nieden,U., Hirt,H. and Scheel,D. (1997)
Receptor-mediated activation of a MAP kinase in pathogen defense of
plants. Science, 276, 2054±2057.
Mazzoni,C., Zarzov,P., Rambourg,A. and Mann,C. (1993) The SLT2
(MPK1) MAP kinase homolog is involved in polarized cell growth in
Saccharomyces cerevisiae. J. Cell Biol., 123, 1821±1833.
Miller,D.D., de Ruijter,N.C.A., Bisseling,T. and Emons,A.M.C. (1999)
The role of actin in root hair morphogenesis: studies with lipochitooligosaccharide
as a growth stimulator and cytochalasin as an actin
perturbing drug. Plant J., 17, 141±154.
Molendijk,A.J., Bischoff,F., Rajendrakumar,C.S.V., Friml,J., Braun,M.,
Gilroy,S. and Palme,K. (2001) Arabidopsis thaliana Rop GTPases are
localized to tips of root hairs and control polar growth. EMBO J., 20,
2779±2788.
Munnik,T., Ligterink,W., Meskiene,I., Calderini,O., Beyerly,J.,
Musgrave,A. and Hirt,H. (1999) Distinct osmosensing protein kinase
pathways are involved in signaling moderate and severe hyperosmotic
stress. Plant J., 20, 381±388.
NuÈhse,T., Peck,S.C., Hirt,H. and Boller,T. (2000) Microbial elicitors
induce activation and dual phosphorylation of the Arabidopsis MAP
kinase AtMPK6. J. Biol. Chem., 275, 7521±7526.
Ovecka,M., BalusÏka,F., NadubinskaÂ,M. and Volkmann,D. (2000)
Actomyosin and exocytosis inhibitors alter root hair morphology in
Poa annua L. Biologia, 55, 105±114.
Robinson,M.J. and Cobb,M.H. (1997) Mitogen-activated protein kinase
pathways. Curr. Opin. Cell Biol., 9, 180±186.
Robinson,M.J., Harkins,P.C., Zhang,J., Baer,R., Haycock,J.W.,
Cobb,M.H. and Goldsmith,E.J. (1996) Mutation of position 52
in ERK2 creates a nonproductive binding mode for adenosine
5¢-triphosphate. Biochemistry, 35, 5641±5646.
Rousseau,S., Houle,F., Landry,J. and Huot,J. (1997) p38 MAP kinase
activation by vascular endothelial growth factor mediates actin
reorganization and cell migration in human endothelial cells.
Oncogene, 15, 2169±2177.
SÏamaj,J., Braun,M., BalusÏka,F., Ensikat,H.J., Tsumuraya,Y. and
Volkmann,D. (1999) Speci®c localization of arabinogalactan±protein
epitopes at the surface of maize root hairs. Plant Cell Phys., 40,
874±883.
Satiat-Jeunemaitre,B., Cole,L., Bourett,T., Howard,R. and Hawes,C.
(1996) Brefeldin A effects in plant and fungal cells: something new
about vesicle traf®cking? J. Microsc., 181, 162±177.
SchaÈfer,C., Ross,S.E., Bragado,M.J., Groblewski,G.E., Ernst,S.A. and
Williams,J.A. (1998) A role for the p38 mitogen-activated protein
kinase/Hsp27 pathway in cholecystokinin-induced changes in the actin
cytoskeleton in rat pancreatic acini. J. Biol. Chem., 273, 24173±24180.
Seo,S., Okamoto,M., Seto,H., Ishizuka,K., Sano,H. and Ohashi,Y. (1995)
Tobacco MAP kinase: a possible mediator in wound signal
transduction pathways. Science, 270, 1988±1992.
Shaw,S.L., Dumais,J. and Long,S.R. (2000) Cell surface expansion in
polarly growing root hairs of Medicago truncatula. Plant Physiol.,
124, 959±969.
Sherrier,D.J. and VandenBosch,K.A. (1994) Secretion of cell wall
polysaccharides in Vicia root hairs. Plant J., 5, 185±195.
Staiger,C.J. (2000) Signaling to the actin cytoskeleton in plants. Annu.
Rev. Plant Physiol. Plant Mol. Biol., 51, 257±288.
Whitmarsh,A.J. and Davis,R.J. (1998) Structural organization of MAPkinase
signaling modules by scaffold proteins in yeast and mammals.
Trends Biochem. Sci., 23, 481±485.
Wilson,C., Pfosser,M., Jonak,C., Hirt,H., Heberle-Bors,E. and
Vicente,O. (1998) Evidence for the activation of a MAP kinase
upon phosphate-induced cell cycle re-entry in tobacco cells. Physiol.
Plant., 102, 532±538.
Woodring,P.J., Hunter,T. and Wang,J.Y.J. (2001) Inhibition of c-Abl
tyrosine kinase activity by ®lamentous actin. J. Biol. Chem., 276,
27104±27110.
Zarzov,P., Mazzoni,C. and Mann,C. (1996) The SLT2 (MPK1) MAP
kinase is activated during periods of polarized cell growth in yeast.
EMBO J., 15, 83±91.
Zhang,S. and Klessig,D.F. (1998) The tobacco wounding-activated
mitogen-activated protein kinase is encoded by SIPK. Proc. Natl
Acad. Sci. USA, 95, 7225±7230.
Received January 11, 2002; revised March 26, 2002;
accepted May 14, 2002
3306

Cell Biology International 27 (2003) 257–259
Short communication
Involvement of MAP kinase SIMK and actin cytoskeleton in the
regulation of root hair tip growth
Jozef Samaj a,e , Miroslav Ovecka a,b,c* , Andrej Hlavacka b , Fatma Lecourieux a ,
Irute Meskiene a , Irene Lichtscheidl d , Peter Lenart a ,Ján Salaj e , Dieter Volkmann b ,
Laszlo Bögre f , Frantisek Baluska b,c , Heribert Hirt a
a Institute of Microbiology and Genetics, Vienna Biocenter, Dr. Bohrgasse 9, 1030 Vienna, Austria
b Institute of Botany, University of Bonn, Kirschallee 1, 53115 Bonn, Germany
c Institute of Botany, Slovak Academy of Sciences, Dúbravská cesta 14, SK – 842 23 Bratislava, Slovak Republic
d Institute of Ecology, University of Vienna, Althanstrasse 14, 1091 Vienna, Austria
e Institute of Plant Genetics and Biotechnology, Slovak Academy of Sciences, Akademická 2, P. O. Box 39 A, SK – 950 07 Nitra, Slovak Republic
f School of Biological Sciences, Royal Holloway, University of London, Egham, TW20 0EX London, UK
Accepted 10 October 2002
Cell
Biology
International
www.elsevier.com/locate/jnlabr/ycbir
1. Introduction
Mitogen-activated protein kinases (MAPKs), a
specific class of serine/threonine protein kinases, are
involved in controlling many cellular functions in all
eukaryotes. Signaling through MAPK cascades is involved
in cell division, differentiation, and stress sensing.
Recently, the stress induced MAPK (SIMK) (Munnik
et al., 1999) and its upstream activator SIMKK (Kiegerl
et al., 2000) have been characterized in Medicago sativa
L. and shown to be inducible by osmotic stress and
various fungal elicitors (Cardinale et al., 2000). In
different plant species dense F-actin meshworks at the
tip of root hairs were observed by immunolabeling
with actin antibodies or in vivo using GFP fused to
the F-actin binding domain of talin (Baluška and
Volkmann, 2002; Baluška et al., 2000a; Braun et al.,
1999). As MAPKs are involved in stress signaling to the
actin cytoskeleton in yeast and animals, we have analyzed
the function of the stress-activated alfalfa MAP
kinase SIMK in root hairs.
2. Results and discussion
In situ hybridization with an SIMK anti-sense probe
revealed that SIMK was strongly expressed in alfalfa
root hairs (Fig. 1). Previously, we have shown that
* Corresponding author. Tel.: +421-2-59426-102;
fax: +421-2-5477-1948.
E-mail address: botuove@savba.sk (M. Ovecka).
Fig. 1. In situ hybridization on alfalfa root sections with SIMK
anti-sense (A) and sense (B, negative control) probes. (A) Superficial
section through root showing that SIMK mRNA (appears as a specific
purple-blue colour) is concentrated to root hairs (indicated by arrows)
emerging along root body surface when section was hybridized with
anti-sense probe. Less intense labeling could also be detected in root
epidermis. (B) Similar superficial section showing no specific labeling
of root section hybridized with the sense probe. Root hairs are
indicated by arrows. Bar=70 µm.
SIMK protein is predominantly localized to nuclei in
meristematic cells of root apices (Baluška et al., 2000b).
SIMK is also found in nuclei of elongating epidermal
root cells. However, during bulge and root hair formation
SIMK is not only polarly relocated from nuclei
towards bulging domains of trichoblasts and tips of
growing root hairs, but it is also activated and located at
root hair tips in an active form. In trichoblasts, SIMK
was located to peripheral spots predicting root hair
outgrowth. In growing root hairs, SIMK was found to
accumulate within root hair tips and in spot-like structures
in the root hair tube. The selective enrichment of
active SIMK in tips of emerging root hairs coincides
1065-6995/03/$ - see front matter 2003 Elsevier Science Ltd. All rights reserved.
doi:10.1016/S1065-6995(02)00344-X

258
J. Samaj et al. / Cell Biology International 27 (2003) 257–259
A
B
cortical
cytoplasm
(A)
nucleus
38 µm (±5 µm)
from the tip(B)
basal cytoplasmic
region
10 µm (C)
sub-apex
10 µm
(D)
apex
7 µm
(E)
C
D
maen fluorescence intensity (AU±SEM)
160
140
120
100
80
60
40
20
0
A B C D E
compartments of root hairs
Fig. 2. A typical example of profile distribution of SIMK labeling in growing root hairs. (A) Immunodetection of SIMK in apical portion of root
hair. Majority of SIMK labeling is visible in apical and sub-apical zones of root hair. Outlined nucleus and line of profile measurement are indicated.
(B) Cytological zonation of growing root hair with indicated scale. (C) The results of relative fluorescence intensities of individual pixels along the
hair horizontal axis (represented by line in A) displayed as the SIMK-level profile diagram with indicated cytological zones of root hair. (D) The
results of mean relative fluorescence intensities of individual root hair zones displayed as grey histogram. n10, Growing root hairs.
with dynamic F-actin meshworks, essential in the onset
of root hair tip growth (Baluška and Volkmann, 2002;
Baluška et al., 2000a; Braun et al., 1999; Miller et al.,
1999). The pattern of SIMK distribution in root hairs
confirmed that it was tip-focused and this cytoarchitectural
gradient was documented by semi-quantitative
fluorescence intensity measurements (Fig. 2).
Tip-focused localization of SIMK disappeared after
treatment with the latrunculin B and resulted in the
relocation of SIMK to the nucleus. Conversely, upon

J. Samaj et al. / Cell Biology International 27 (2003) 257–259 259
actin stabilization with jasplakinolide, an inducer of
actin polymerization, SIMK co-localized with thick
actin cables in the cytoplasm. Importantly, latrunculin B
and jasplakinolide were both found to activate SIMK in
a root-derived cell culture (Samaj et al., 2002). Loss of
tip-focused SIMK and actin was induced by the MAPK
kinase inhibitor UO 126 and resulted in aberrant root
hairs. UO 126 inhibits polarized root hair growth by
changing polar vesicle trafficking and root hair cytoarchitecture.
After UO 126 treatment, arrest of hair
growth and vacuolation of root hair tips were observed
by video enhanced microscopy. Tubular small vacuoles
within the tip inflated to large roundish vacuoles that
oscillated in random motion, and the shape of the apical
dome became swollen. SIMK was distributed evenly in
cytoplasm and nuclei of UO 126 treated hairs without
any preferable accumulation in subcellular compartments
(Samaj et al., 2002). Thus, UO 126 inhibits root
hair tip growth by modifying the cytoarchitecture and
actin-dependent polar vesicle targeting and trafficking.
In contrast to SIMK inhibition, overexpression of gainof-function
SIMK induced rapid tip-growth of root
hairs and could bypass growth inhibition by UO 126.
3. Conclusions
An intact actin cytoskeleton and vesicle trafficking
are necessary for the proper localization of SIMK within
root hair tips. Depolymerization and stabilization of
F-actin activate SIMK, indicating that the MAPK is not
only associated with the actin cytoskeleton but its
activity is also directly affected by altering the dynamics
of F-actin. Inhibition of MAPK activity caused remodeling
and changes in the subcellular distribution of
actin and SIMK, resulting in tip-growth inhibition and
aberrant root hair morphology. On the other hand,
overexpression of gain-of-function SIMK in transgenic
plants resulted in increased root hair formation and
growth. Taken together, our data suggest that SIMK
plays an important role in root hair tip growth linking
MAPK signaling to actin cytoskeleton.
References
Baluška F, Volkmann D. Actin-driven polar growth of plant cells.
Trends Cell Biol 2002;12:14.
Baluška F, Salaj J, Mathur J, Braun M, Jasper F, Šamaj J et al. Root
hair formation: F-actin-dependent tip growth is initiated by local
assembly of profilin-supported F-actin meshworks accumulated
within expansin-enriched bulges. Dev Biol 2000;227:618–32.
Baluška F, Ovecka M, Hirt H. Salt stress- and cell cycle phasedependent
changes in expression and subcellular localisation of the
alfalfa mitogen-activated protein kinase SIMK. Protoplasma 2000;
212:262–7.
Braun M, Baluška F, Von Witsch M, Menzel D. Redistribution of
actin, profilin and phosphatidylinositol-4,5-bisphosphate (PIP 2 )in
growing and maturing root hairs. Planta 1999;209:435–43.
Cardinale F, Jonak C, Ligterink W, Niehaus K, Boller T, Hirt H.
Differential activation of four specific MAPK pathways by distinct
elicitors. J Biol Chem 2000;275:36734–40.
Kiegerl S, Cardinale F, Siligan C, Gross A, Baudouin E, Liwosz A
et al. SIMKK, a mitogen-activated protein kinase (MAPK)
kinase, is a specific activator of the salt stress-induced MAPK,
SIMK. Plant Cell 2000;12:2247–58.
Miller DD, De Ruijter NCA, Bisseling T, Emons AMC. The role
of actin in root hair morphogenesis: studies with lipochitooligosaccharide
as a growth stimulator and cytochalasin as an actin
perturbing drug. Plant J 1999;17:141–54.
Munnik T, Ligterink W, Meskiene I, Calderini O, Beyerly J, Musgrave
A et al. Distinct osmosensing protein kinase pathways are involved
in signaling moderate and severe hyperosmotic stress. Plant J 1999;
20:381–8.
Samaj J, Ovecka M, Hlavacka A, Lecourieux F, Meskiene I,
Lichtscheidl I et al. Involvement of the mitogen-activated protein
kinase SIMK in regulation of root hair tip-growth. EMBO J 2002;
21:3296–306.

Protoplasma (2010) 243:51–62
DOI 10.1007/s00709-009-0055-6
ORIGINAL ARTICLE
Plasmolysis and cell wall deposition in wheat root hairs
under osmotic stress
Michael Volgger & Ingeborg Lang & Miroslav Ovečka &
Irene Lichtscheidl
Received: 13 February 2009 /Accepted: 25 May 2009 /Published online: 17 June 2009
# Springer-Verlag 2009
Abstract We analysed cell wall formation in rapidly growing
root hairs of Triticum aestivum under reduced turgor pressure
by application of iso- and hypertonic mannitol solutions. Our
experimental series revealed an osmotic value of wheat root
hairs of 150 mOsm. In higher concentrations (200–
650 mOsm), exocytosis of wall material and its deposition,
as well as callose synthesis, still occurred, but the elongation
of root hairs was stopped. Even after strong plasmolysis when
the protoplast retreated from the cell wall, deposits of wall
components were observed. Labelling with DiOC 6 (3) and
FM1-43 revealed numerous Hechtian strands that spanned the
plasmolytic space. Interestingly, the Hechtian strands also led
towards the very tip of the root hair suggesting strong
anchoring sites that are readily incorporated into the new cell
wall. Long-term treatments of over 24 h in mannitol solutions
(150–450 mOsm) resulted in reduced growth and
concentration-dependent shortening of root hairs. However,
the formation of new root hairs does occur in all concentrations
used. This reflects the extraordinary potential of wheat
root cells to adapt to environmental stress situations.
Dedicated to Professor Cornelius Lütz on the occasion of his 65th
birthday
M. Volgger : I. Lang (*) : I. Lichtscheidl
Cell Imaging and Ultrastructure Research,
Faculty of Life Sciences, The University of Vienna,
Althanstrasse 14, 1090 Vienna, Austria
e-mail: ingeborg.lang@univie.ac.at
M. Ovečka
Agrobiotechnology Institute, Campus de Arrosadia,
Mutilva Baja 31192 Navarra, Spain
M. Ovečka
Institute of Botany, Slovak Academy of Sciences,
Dubravska cesta 14, 84523 Bratislava, Slovakia
Keywords Cell wall . Mannitol osmotic stress .
Plasmolysis/Hechtian strands . Root hairs . Tip growth .
Triticum aestivum
Abbreviations
CESA6 Cellulose synthase 6 protein
DCB 2,6-Dichlorobenonitrile
MAPK Mitogen activated protein kinase
YFP Yellow fluorescent protein
Introduction
Root hairs are tubular elongations of the rhizodermis (root
epidermis) and serve as contact between plant and soil to
improve water and mineral uptake. Their unidirectional
growth is facilitated by the deposition of cell wall material
at a very limited area which allows for targeted observation
of the process and has made root hairs a model for analysis
of cell wall formation (Hepler et al. 2001; Baluška et al.
2003; Šamaj et al. 2004). At the same time, their important
role for plant nutrition makes root hairs and their response
to soil conditions especially interesting for crop yield.
Cellular organisation of root hairs and cell wall formation
A growing root hair has a tubular shape and can be divided into
three zones (Volkmann 1984): the tip zone with a clear apical
zone and subapical zone, the vacuolation zone and the basal
zone (Fig. 1). (1) The tip zone consists of a polar
accumulation of cytoplasm and small vesicles (Galway et al.
1997; Galway,2000). Cell elongation occurs at the domeshaped
front of this apical zone (“tip growth”) (Hepler et al.
2001; Sieberer et al. 2005). Golgi vesicles that contain

52 M. Volgger et al.
Fig. 1 Zones of a growing
root hair: basal zone close to the
rhizodermis (a), vacuolation
zone (b), tip zone with
clear zone at the very front (c).
Bar 10 μm
pectins, hemicelluloses and polygalacturonic acid (McNeil
et al. 1984; VarnerandLin1989) fuse with the existing
plasma membrane and deposit their content towards the
outside of the growing cell by exocytosis. At the same time,
excess membrane material is recycled by endocytotic
processes (Ovečka et al. 2005). In addition, some callose
and cellulose depositions were observed after staining with
aniline blue (callose) and calco fluor white (cellulose) (Strobl
and Lichtscheidl 2004). The subapex contains larger cell
organelles and compartments like endoplasmatic reticulum
(ER), mitochondria and Golgi stacks. The cell walls on the
flanks of the root hairs are thicker and less plastic than at the
dome. At the sides, cellulose microfibrils form primary and
then secondary cellulose textures that support and restrain the
growing root hair (Schröter and Sievers 1971; Schnepf et al.
1986). (2) In the vacuolation zone, the central vacuole fills
most of the cell and, except for some cytoplasmic strands, the
cytoplasm is pushed towards the side walls. Here, organelles
show a rotational streaming. It is postulated that turgor
pressure of the vacuole is the driving force for cell expansion
in general and tip growth in particular. (3) The basal zone of
the root hair lies close to the rhizodermis. The vacuole
completely fills also this part of the root hair, and the thin
layer of cytoplasm is pushed towards the cell wall.
Cell wall formation under reduced turgor
Turgor pressure of the vacuole can be reduced under osmotic
stress conditions, as induced by freezing, drought or soil
salinity. Are tip growth and cell wall formation still possible in
a turgorless state? Here, this situation is simulated by
plasmolysis (de Vries 1877; Oparka1994). The hypertonic
(plasmolytic) solution osmotically extracts water from the
vacuole up to a point of complete turgor loss and, depending
on the concentration of the solution, can even cause the
detachment of the living protoplast from the cell wall
(Schnepf et al. 1986; Lang-Pauluzzi 2000). Thin, threadlike
connections have been observed between the protoplast and
the cell wall of plasmolysed cells. These connections are
named after Hecht (1912) and are referred to as Hechtian
strands (Sitte 1963; Schnepf et al. 1986; Oparkaetal.1996;
Bachewich and Heath 1997; Lang-Pauluzzi and Gunning
2000;Langetal.2004;DeBoltetal.2007). At the cell wall of
plasmolysed plant cells, Hechtian strands pass into a cortical
network, the Hechtian reticulation (Pont-Lezica et al. 1993)
that reminds of cortical endoplasmatic reticulum. Studies by
Oparka et al. (1993, 1996) on plasmolysed onion epidermal
cells demonstrated that, although the cortical ER remained
strongly attached to the cell wall, it is nonetheless bounded by
the plasma membrane. Together, Hechtian strands and
Hechtian reticulum conserve the surface area of the plasma
membrane in plasmolysis/deplasmolysis cycles (Oparka et al.
1994). Hechtian strands may form the basis for freezing
tolerance in cold-hardened plants: as the temperature of the
cell is lowered, ice grows outside the cell by the withdrawal
of water from inside. Cold-hardened cells have been shown
to form more Hechtian strands than non-hardened cells
(Johnson-Flanagan and Singh 1986; Singh et al. 1987; Buer
et al. 2000). The authors conclude that membrane conservation
by Hechtian strands is a means to successfully retrieve
membrane material during expansion in deplasmolysis.
In the present study, the impact of plasmolytic solutions
on cell wall formation was tested in living root hairs of the
crop plant Triticum aestivum. The adaptation to osmotic
stress situations and the ability to form new root hairs under
these conditions were analysed by long-term culture in
plasmolytic solutions. We provide clear evidences for
continuation of the cell wall formation process in root hairs
under turgorless conditions, although no further tip growth
is possible. Together with formation of new root hairs in
higher osmolarity, it shed more light to the adaptation
abilities of tip-growing root hairs.
Materials and methods
Plant material
Living roots and root hairs of T. aestivum (Poaceae) were
prepared according to the protocol of Ovečka et al. (2005).

Plasmolysis and cell wall deposition in root hairs 53
In brief, seeds were synchronised in distilled water for
2 days at 4°C and then germinated in Petri dishes on wet
filter paper for 1 day at 24°C under continuous light
conditions. Three- to 4-day-old seedlings were transferred
into micro-chambers between a cover slip and a glass slide.
Six layers of Parafilm “M” (Pechiney Plastic Packaging,
Chicago, IL, USA) acted as a spacer between slide and
cover slip. In vertical orientation, the seedlings grew in
sterile glass cuvettes for another 24–30 h under continuous
light. As culture medium, we used a phosphate buffer, pH
6.2, 25 mOsm. The medium just reached the open lower
edge of the micro-chambers allowing for free exchange of
the medium between chambers and the cuvette. During the
cultivation period, the seedlings were exposed to constant
light at 22°C. To prevent damage or unintentional changes
of growth, it was very important to protect the very
sensitive root hairs against mechanical stress. The custommade
growth chambers enabled us to take the roots to the
microscope stage and analyse them without further transfer
or manipulation. Prior to every experiment, the regular and
constant growth of the selected cells was controlled. To
assure standard conditions, we observed root hairs during
the most stable phase of growth at a length of 100–300 μm.
For adaptation studies of whole roots, the seedlings were
prepared as described above and grown under the respective
osmotic stress conditions for up to 30 h. Data were
taken at 24 h after transfer. Lengths of whole roots were
measured before and after transfer to the osmotic solution.
During the time of observation, wheat seedlings formed one
primary root and two lateral roots. Total root length was
calculated by the addition of the lengths of the primary root
plus lateral roots. Out of three seedlings per concentration,
a mean value was determined reflecting the increment.
Editing of data and graphic design was done in Microsoft
Access and Excel (Microsoft Office 2003). Statistics were
performed on SPSS[R] 10.0. Correlation between length of
root hairs and osmotic value of growth media was tested by
Spearman rank correlation for standard significance level of
[alpha]=0.01%.
Osmotic solutions
Osmotic stress of the plants was induced by exposure to
iso- and hypertonic solutions of glucose (Merck, Germany)
and D-mannitol (Merck, Germany) in concentrations of 100
to 650 mOsm. Under the microscope, the culture medium
of the micro-chamber was carefully removed with a strip of
filter paper and replaced by the osmotic solution with a
micropipette (chamber perfusion). For long-term experiments,
the buffer in the cuvette was replaced by the osmotic
solution. After germination and transfer to the microchambers,
the seedlings were grown in the osmotic medium
for up to 30 h. Osmotic values of solutions were measured in a
Micro-Osmometer (Model 3MO plus, Advanced Instruments,
Massachusetts, USA).
The osmotic value of root cells and root hairs was
determined microscopically and coincides with the state of
incipient plasmolysis (or limiting plasmolysis; Oparka 1994)
equalling the respective concentration of gradient osmotic
solutions.
Autofluorescence and fluorescent markers
The deposition of new cell wall material, mainly callose,
was visualised by its blue autofluorescence after UV
excitation. Selective staining of callose was performed by
chamber perfusion with 1% aniline blue at a concentration
in buffer or osmotic solution, respectively. After UV
excitation, aniline blue gave a yellow fluorescence.
The membrane potential dye DiOC 6 (3) (3,3′-dihexyloxacarbocyanine
iodide) (Molecular Probes, USA) was used
to stain the Hechtian strands, Hechtian reticulum and ER of
plasmolysed root hairs. DiOC 6 (3) was applied by perfusion
of the micro-chamber for 5 min, just prior to observation.
We used a concentration of 5 μg/ml dissolved in the
respective osmotic solution and rinsed off excess dye with
the same osmotic solution. DiOC 6 (3) has been successfully
used to stain Hechtian strands in onion and Tradescantia
cells (Oparka et al. 1994; Lang-Pauluzzi and Gunning
2000; Lang et al. 2004).
Alternatively, roots were exposed to a membrane selective
non-permeable styryl dye, FM1-43 (N-(3-triethylammoniumpropyl)-4-(4-[dibutylamino]styryl)pyridinium
dibromide)
(Molecular Probes). The dye has been widely used for
observing plasma membrane recycling (Betz et al. 1996;
Emans et al. 2002; Bolte et al. 2004; Ovečka et al. 2005). It
was applied by perfusion for 4 min right before observation.
We used a final concentration of 8 μM FM1-43 in buffer or
osmotic solution. With an excitation wavelength of 488 nm,
the emission maximum of FM1-43 lies at 600 nm.
Microscopy
A Labophot 2 (Nikon) with epi-fluorescence illumination
was used for conventional light microscopy. Autofluorescence
of cell wall components and aniline blue labelling
was detected after UV excitation in combination with a UV-
2A/DM400 (Nikon) emission filter set.
Furthermore, living root hairs were analysed by confocal
laser scanning microscopy (Leica DMIRE2) with an objective
×63, NA 1.32 (Leica). For DiOC 6 (3) and FM1-43 labelling,
the 488 nm excitation wavelength from an argon laser was
selected. Step sizes for z-series varied from 0.25 to 0.8 μm.
Stacks of images and maximum projections were generated
with the software associated to the laser scanning microscope
(Leica Confocal Software). Images were taken in the

54 M. Volgger et al.
fluorescence and transmission mode simultaneously to clarify
the location of membranes within plasmolysed root hairs.
Overviews of seedlings and whole roots were taken with a
Coolpix 990 (Nikon) attached to a stereo microscope (Nikon).
Results
Root hairs are formed by bulging from root epidermal cells.
They grow up to a length of 500 to 1,000 μm (Fig. 2a) in
about 13 h. During bulge formation, growth is slow with
0.1 to 0.2 μm/min. At a length of 5 to 35 μm, growth
speeds up continuously to 1 μm/min. This velocity is
maintained until the middle of the total growth phase (about
7 h). Thereafter, growth continuously slows down till the
end of the growth phase (Fig. 2 b, c); figures were taken in
10 min intervals, the light source was switched off between
the takes. The resulting growth curve was confirmed by at
least 100 random samples of root hairs that were not
stressed by long-term microscopic observation.
When root hairs cease elongation, the polar distribution
is lost, and the vacuole progresses into the tip zone. The
organelles distribute regularly over the whole length of the
tube, and also large organelles reach up into the clear zone.
The velocity of organelles, however, is maintained.
Isotonic medium causes growth stop
Up to a concentration of 150 mOsm mannitol, neither
changes in growth nor cytoarchitecture of root hairs could
Fig. 2 a–c Root hair elongation
under normal growth conditions
(n

Plasmolysis and cell wall deposition in root hairs 55
be detected (Fig. 3a). At this concentration of 150 mOsm,
we observed two different reactions: some root hairs are not
affected by the osmotic medium and maintain normal tip
growth, whereas others stop elongation after transfer into
the osmotic medium. In both cases, plasmolysis does not
occur. We therefore conclude that the osmotic value of T.
aestivum root hairs is around 150 mOsm; cells in this
situation are in a turgorless state.
Deposition of new cell wall in the tip continues and the
polar organisation of the cytoplasm remains. After only
short treatment with isotonic medium, cell elongation can
be resumed after transfer back to original growth medium.
When we kept the cells, however, for longer periods in
isotonic buffer, growth could not be renewed despite the
continuous asymmetric cellular organisation.
Continuous deposition of wall material during plasmolysis
In mildly hypertonic solutions of 200–350 mOsm, water
loss from the vacuole into the hypertonic medium occurs,
and root hairs stop growing. The vacuole shrinks, and the
protoplast retreats from the cell wall (plasmolysis). The
retraction of the protoplast starts from the tip but is very
slow. Cell wall material is still discharged and deposited in
the emptying tip of root hairs where it builds up a thick
layer (Fig. 3b, c). The new wall material shows a dispersed
and inhomogeneous accumulation of cell wall material and
completely fills the space between the protoplast and the
cell wall (Fig 3c). This state of mild plasmolysis can be
maintained for several hours. During this time, neither
modifications in the cytoarchitecture nor in the mode or the
velocity of cytoplasmic streaming were observed.
The deposited wall material consists mainly of callose
which can be stained with aniline blue (Fig. 4a). The newly
deposited wall is also clearly visible by its autofluorescence
under UV excitation (Fig. 4b, c).
Due to further loss of water into the hypertonic medium of
strong concentrations over 350 mOsm, plasmolysis proper
occurs. The vacuole becomes so small that the protoplast
retreats and detaches from the cell wall until an isotonic
equilibrium is reached (Fig. 3d). It draws back mainly from the
tip, but we observed also slight detachments along the flanks
of the hair at the vacuolation zone. Even during protoplast
retraction, cell wall material is excreted, but no persistent cell
wall can be built up like in lower concentrations. Instead,
bands of new wall material are visible along the cell wall of
the empty tip (Fig. 3d). Interestingly, this process takes almost
1 h in root hairs which is much longer than plasmolysis needs
under such conditions in mature cells from stems and leaves
of T. aestivum (data not shown). At this stage, the protoplast is
arrested, and new cell wall material is deposited at the outside
of the retracted protoplast (Fig. 5e). This new cell wall
appears solid and dense in contrast to the cell wall built in
lower osmotic concentrations. Its formation takes 45 min.
A successive, stronger plasmolysis at 650 mOsm causes
renewed, further contraction of the protoplast. Thereby, the
Fig. 3 a–d Osmotic stress after
direct transfer of a root hair
into plasmolytic solutions.
a 150 mOsm mannitol. Two
populations of root hairs are
observed. One shows no
affection, the other performs a
short stop and resumption of
growth. b, c 250–350 mOsm
mannitol. The cells are in a
turgorless state but no
detachment of the protoplast
from the cell wall occurs. The
continuing deposition of wall
material results in wall
thickening at the tip (arrows).
d Over 400 mOsm mannitol.
The protoplast detaches from the
cell wall and contracts until an
osmotic equilibrium is reached;
deposition of wall material is
continued (arrow heads); the
polar organisation of the cell
is maintained. Asterisks mark
the plasmolysed protoplast.
Bar 10 μm

56 M. Volgger et al.
The plasma membrane remains connected to the cell wall
after plasmolysis
During strong plasmolysis (350 mOsm and above),
Hechtian strands are formed (Fig. 5). DiOC 6 (3), a
membrane-selective dye, was used to visualise the thin
Hechtian strands and the Hechtian network close to the
cell wall. Figure 5a, b shows this membranous network at
different focal planes. It is evident that there is no
symmetric pattern of the strands. The Hechtian strands
are anchored to the shank of the root hair and also reach
towards the tip (Fig. 5 a, b, d). In bright field mode, the
Hechtian strands are merely visible (Fig. 5c). Alternatively
to DiOC 6 (3), these thin strands are brightly stained with
FM1-43, a membrane-selective styryl dye (Fig. 5d). FM1-
43 is only fluorescent when incorporated into the plasma
membrane and has been widely used as endocytosis
marker (Emans et al. 2002; Bolte et al. 2004; Ovečka
et al. 2005). With FM1-43, we observed Hechtian strands
emerging from the main protoplast as well as from subprotoplasts.
The strands are anchored to the flanks of the
root hair wall, as described for DiOC 6 (3), and emerged
also from the clear zone at the very tip of the root hair
(Fig. 5d). Thus, strong attachment sites for Hechtian
strands have to be incorporated already within the very
young, newly deposited cell wall.
Root hairs adapt to osmotic stress
Fig. 4 a–c Root hair tips under the fluorescence microscope. a–c Cell
wall deposition at 300 mOsm mannitol; a callose staining with aniline
blue (arrow). b callose/cellulose autofluorescence after UV excitation
in root hair tip showing dispersed material and a bright band of newly
formed cell wall (arrow). c Superimposing autofluorescence of wall
material after UV excitation and brightfield shows the plasmolysed
protoplast. Asterisks mark the plasmolysed protoplast. Bar 10 μm
cell wall that had been built at 400 mOsm becomes clearly
visible in the emptying plasmolytic space (Fig. 5f).
The intracellular organisation of the cytoplasm is maintained
at first. Only after 2 h in strong hypertonic solutions
over 350 mOsm, we observed a reorganisation of the
cytoarchitecture: the reverse fountain streaming is changed
into circulation streaming, and the clear zone at the tip is
reduced, however maintained. The position of the nucleus as
well as the velocity of organelle movement is constant.
In another set of experiments, we wanted to test the
adaptation of root hairs to osmotic stress conditions. For
that purpose, wheat roots were grown for 24 h in glass
cuvettes containing the osmotic solutions with concentrations
of 150 to 450 mOsm. Growth of root hairs and the
ability to form new root hairs under long-term osmotic
stress were tested as well as the development of whole
roots.
Figure 6 shows a significant reduction of the total
number of root hairs built (Fig. 6 a–d, dark field) and length
of root hairs (Fig. 6e–h). Under control conditions and up to
isotonic 150 mOsm, root hair development appears normal
(Fig. 6a, e). At 250 mOsm, less and shorter root hairs are
built than at lower concentrations (Fig. 6b, f). Interestingly,
even after 24 h, newly formed root hairs appear in osmotic
solutions of 350 and 450 mOsm mannitol, when normal cells
react with strong plasmolysis. Under these long-term osmotic
conditions, root hairs do not grow beyond a length of 50 to
100 μm, but their cellular organisation is the same as in
control hairs (Fig. 6e–g). Only at 450 mOsm, the clear zone
at the tip is not developed properly (Fig. 6h). Plasmolysis of
these “hardened” root hairs is only possible with concentrations
that are 200 mOsm stronger than during cultivation
(data not shown). Even so, they show only mild plasmolysis

Plasmolysis and cell wall deposition in root hairs 57
Fig. 5 a–f Plasma membrane–
cell wall connections.
a, b Fluorescence micrograph of
clearly visible Hechtian strands
(arrow heads) after labelling
with DiOC 6 (3), two different
focal planes. c Corresponding
bright field to a and b.
d Hechtian strands (arrow head)
towards the very tip after
labelling with FM1-43.
e Plasmolysis in 400 mOsm
mannitol leads to the formation
of a plasmolytic space (ps1),
newly formed cell wall at the tip
of the protoplast (arrow).
f Further plasmolysis of the
cell in e in 650 mOsm mannitol,
the wall built in 400 mOsm
mannitol remains in place
(arrow); plasmolytic space
2(ps2) is formed. Asterisks
mark the plasmolysed
protoplast. Bar 10 μm
and a thickening of the cell wall at the tip, suggesting a
change of internal osmotic value.
Root hairs reach up to 1,000 μm in length under control
conditions and still up to 700 μm in isotonic 150 mOsm
mannitol. Interestingly, the decrease in length is significant
(r=−0.976, P≤1‰) but not linear: at 250 mOsm, root hairs
grow less than at 350 mOsm. From 350 mOsm to
450 mOsm, the length of root hairs gradually diminishes
until root hair formation is finally stopped (Fig. 7a). Up to
100 mOsm above the isotonic concentration of wheat root
hairs, cells apparently struggle to adapt, resulting in a
diminished growth. Thereafter, elongation is again possible
before gradual reduction occurs. We interpret this observation
with the possibility of the cells to adapt up to a certain
osmotic limit before plasmolysis occurs. Interestingly, the
growth rate of stressed root hairs is identical to control cells
(Fig. 2c) in all osmotic solutions tested. For wheat root
hairs, we measured a maximum growth rate of 1 μm/min.
Similarly to root hairs, the lengths of whole roots also
gradually shorten with increasing osmotic stress: in
150 mOsm mannitol, roots grow up to 40 μm whereas in
550 mOsm, they gain only 19 μm over the same period of
30 h (Fig. 7b).
Discussion
Cell wall formation of root hairs of T. aestivum was
investigated under two different osmotic situations: (1)
Direct transfer of the cells to different osmotic solutions
allowed us to analyse the reactions on the cellular level:
growth continuity, cell wall deposition and cytoarchitecture.
(2) Cells kept in osmotic solutions for 24 h focussed on the
environmental situation: the potential to form new root
hairs within this time and the adaptation of the cells to
elevated osmotic conditions.

58 M. Volgger et al.
Fig. 6 a–h After 24 h of incubation: adaptation of root hairs to
plasmolytic conditions. The left column shows whole roots and root
hairs in increasing concentrations of mannitol, dark field mode; the
right column shows single root hairs under these conditions in higher
magnification, bright field mode. Bar left column 30 μm; bar right
column 10 μm
Organisation of the cytoplasm and deposition
of wall material
In root hairs of Tradescantia fluminensis, Schröter and
Sievers (1971) observed differences in cellular organisation
after 3 h in 0.05 molar glucose solutions and a cessation of
growth in existing hairs.
Our results of wheat root hairs show that the cytoarchitecture
is not influenced dramatically during osmotic
changes; even in plasmolysed root hairs, the clear zone
and polar organisation of the cells are maintained for a long
time independent of the osmotic value of the medium. In
addition, the deposition of new cell wall is continuous
during all stages of plasmolysis. It is secreted into the
emerging plasmolytic space between the cell wall and the
protoplast. Obviously, the tip-focussed gradient of exocytotic
vesicles and the continuing organelle motility allow
for a constant supply of wall components and for continued
cell wall synthesis. In plants, disrupted cellulose synthase
may form callose (Brett 2000). Depending of the speed of
plasmolysis, wall depositions appear dispersed or as solid
rings. They remain in place and become especially visible
after renewed plasmolysis.
This new cell wall can be observed by autofluorescence
or after callose staining with aniline blue just exterior of the
plasma membrane, and it appears brighter than within the
plasmolytic space. Electron micrographs of cell walls built
under osmotic stress show a rather loose consistency and
confirm the deposition of mainly pectins and hemicelluloses
in moss protonemata (Schnepf et al. 1986) and in root
hairs of T. fluminensis (Schröter and Sievers 1971). In root
hairs of Zea mays and T. aestivum, also callose depositions
have been observed after turgor reduction (Lerch 1960). In
Tradescantia virginiana leaf epidermal cells, we also found
striking callose depositions at the plasmolysed protoplast,
the inner side of the cell wall (Lang et al. 2004).
Plasmolysis seems to be a membrane-driven process which
still takes place after disruption of actin microfilaments with
cytochalasin B (M Volgger, data not shown) and, in onion
epidermis cells, plasmolysis and Hechtian strand formation was
still observed after cytoskeleton elements had been depolymerized
(Lang-Pauluzzi and Gunning 2000). In osmotically
stressed cells, additional osmocytotic vesicles are formed to
reduce membrane surface area of shrinking protoplasts
(Oparka et al. 1990; Oparkaetal.1996). Interestingly, these
authors demonstrated that osmocytotic vesicles were not
reused in deplasmolysis of onion epidermal cells.
Formation and attachment of Hechtian strands
In strong plasmolysis, Hechtian strands form between the
plasma membrane and the cell wall of plant tissue (Hecht 1912;
Attree and Sheffield 1985;Oparka1994; Lang-Pauluzzi 2000;
Lang et al. 2004), and we show the strands here after FM1-34
and DiOC 6 (3) staining in a tip growing system, i.e. root hairs.
Hechtian strands reached right towards the newly built wall at
the clear zone, suggesting that the anchoring sites are readily
incorporated at the very tip. The strands clearly form without
plasmodesmata (Pont-Lezica et al. 1993) and thus, the
initiation sites of Hechtian strands within the cell wall have

Plasmolysis and cell wall deposition in root hairs 59
Fig. 7 a–b Diagrams of root
hair (a) and whole root (b)
lengths under osmotic stress
conditions. a Up to a
concentration of 250 mOsm
mannitol, length of root hairs
diminishes; at 350 mOsm, root
hairs grow significantly longer
than in 250 mOsm before they
are shortening again in very
high concentrations of mannitol.
b Whole roots continuously
shorten with increasing osmotic
stress. Growth rates are identical
to controls
been fiercely discussed. Some authors suggested integrins
known from animal cells (Ruoslahti and Pierschbacher 1987;
Canut et al. 1998; Kiba et al. 1998; Barthou et al. 1999;
Hostetter 2000), others focussed on wall-associated kinases as
attachment sites (Kohorn 2001) or arabinogalactans (Gouget
et al. 2006). A comprehensive review on the subject is given
by (Baluška et al. 2003). From our studies on tip-growing root
hairs, it is evident that the strong anchorage of Hechtian
strands and the Hechtian reticulum is initiating in the primary
wall. Indeed, very recent studies using genetic approaches on
Arabidopsis thaliana revealed the organisation of cellulose
synthase complexes in the primary cell wall (Desprez et al.
2007; Persson et al. 2007; Wangetal.2008a). Furthermore,
inhibition of cellulose synthase complexes in combination
with their visualisation in A. thaliana mutants finally shed
some light on the postulated connection of cellulose microfibrils
and cortical microtubules in living cells (Paredez et al.
2006). After plasmolysis, YFP:CESA6-labelled cellulose
synthase complexes remained in the plant cell wall (DeBolt
et al. 2007). This would be consistent with our model
whereby Hechtian strands are tethered to the cellulose
microfibril array via CESA rosettes in the plasma membrane
(Lang et al. 2004). Recently, Paredez et al. (2006) elegantly
showed the movement of cellulose rosettes during wall
deposition, and hence, the postulation of motile synthase
rosettes as anchoring sites for Hechtian strands is no longer
suitable unless during plasmolysis, the movement of rosettes
is blocked. Furthermore, the inhibition of cellulose synthesis
with the drugs 2,6-dichlorobenonitrile (DCB) or isoxaben
prior to plasmolysis resulted in YFP:CESA6-labelled complexes
localised to the plasmolysed protoplast and not to
the cell wall. DCB and isoxaben pre-treatment did not affect
the formation of Hechtian stands, suggesting that their
formation is independent of CESA rosettes (DeBolt et al.
2007). Therefore, it appears unlikely that CESA is solely
responsible for Hechtian strand adhesion to the cell wall, and

60 M. Volgger et al.
other candidates (as mentioned above) will have to be
re-considered.
The Hechtian reticulum may play a role in deplasmolysis
In fact, not only punctuate Hechtian strand adhesion sites
remain attached to the inner surface of the cell wall in
plasmolysed plant cells. A network that reminds of cortical
ER, the Hechtian reticulum, was observed in plasmolysed
onion inner epidermal cells after staining with DiOC 6 (Oparka
et al. 1990; Oparka 1994; Lang-Pauluzzi and Gunning 2000).
It is similar to reticulum that lined the inner walls of the root
hairs after staining for plasma membrane and ER. It was
suggested that the same membrane-spanning linkages which
attach the plasma membrane to the wall on its external face
might also anchor the ER to its internal face (Oparka 1994).
Elements of the actin cytoskeleton could be such candidates;
they are closely associated with the plasma membrane and
the cortical ER (Lichtscheidl and Url 1990; Lichtscheidl and
Hepler 1996; Overall et al. 2001; Baluška et al. 2003),
possibly providing a scaffold for the re-affixture or realignment
of the protoplast during deplasmolysis. In the root hairs
that we investigated here, deplasmolysis usually led to
disruption of the cells. Obviously, the newly synthesised
wall at the tip is too weak and misses a “rescue pattern” to
withstand that pressure.
Roots adapt to hypertonic conditions
Exposure of roots to hypertonic media stopped root hair
growth immediately. However, during prolonged treatment,
roots themselves continued to grow, and they even developed
new root hairs in mannitol solutions as high as 450 mOsm. The
adapted roots were shorter, and also the newly formed root
hairs were shorter, although the velocity of their growth and
their polar organisation were identical to the control (Fig. 6).
Their osmotic value was significantly higher; 250 mOsm
corresponds to the onset of proper plasmolysis and detachment
of the protoplast from the cell wall in T. aestivum.
Accordingly, it needed much higher osmotic values to induce
plasmolysis. Obviously, osmotically active substances were
mobilised that help to adjust to the surrounding medium.
From the curve in Fig. 7a, we conclude that such elevation of
the osmotic value of the wheat root hair is triggered by a
plasmolysing concentration of 300 mOsm and above.
A putative role of aquaporins during plasmolysis
Aquaporins may play a key role during osmotic stress
situations, and both types of aquaporins, plasma membrane
intrinsic proteins and tonoplast intrinsic proteins (TIPs),
have been shown to occur in plants (for review, see
Johansson et al. 2000 and/or Tyerman et al. 2002).
Recently, osmotic water permeability of plasma and
vacuolar membranes was elegantly measured in radish
(Raphanus sativus) protoplasts and high aquaporin activity
observed in both the plasma membrane and the tonoplast
(Murai-Hatano and Kuwagata 2007). At present, all
examined plant aquaporins have enhanced the osmotic
water permeability, although aquaporin function, dependent
on osmotic gradients, is a passive element. Thus, the rate of
aquaporin-mediated water transport also depends on other
active transporters and ion channels. However, vacuolar
membranes containing abundant TIPs are effective to
prevent plasmolysis (Hara-Nishimura and Maeshima 2000).
At the protein level, aquaporins are not the only group of
targeted candidates. Several protein kinases are also known
to be involved in both salt and osmotic stress as well as in
wounding and drought, e.g. SIMK and SAMK (Jonak et al.
1996; Meskiene and Hirt 2000; Šamaj et al. 2002; Ovečka
et al. 2008b). Furthermore, plant responses to abiotic
stresses like drought stress, osmotic stress or wounding
are closely related to signalling roles of plant hormones
ABA and jasmonic acid (Denekamp and Smeekens 2003).
Using mutant analysis, many Arabidopsis genes have been
identified that affect the process of tip growth (Parker et al.
2000). One of them is the Arabidopsis sos4 mutant, salt
overly sensitive 4, that was isolated by screening for NaClhypersensitive
growth. The SOS4 gene encodes a pyridoxal
kinase involved in the production of pyridoxal-5-phosphate.
The general phenotype of sos4 is salt hypersensitivity, but
additionally, sos4 roots failed to form root hairs, and
rhizodermis showed only a few bulges (Shi and Zhu
2002). Further analysis of the salt overly sensitive mutants
showed a dose-dependent reduction of root hair density by
salt treatments (Wang et al. 2008b). The authors found that
Na + ,K + and Li + , but neither the closely related Cs + nor
mannitol stress, caused inhibition of root hair development.
By contrast, osmotic stress caused by 200 mM mannitol
increased root hair density and promoted root hair growth
rate in stressed Arabidopsis roots. This suggests that the
inhibitory effects of salt on root hair development were
caused by ion disequilibrium and not by osmotic effects. In
the present study, we therefore deliberately focussed on
plasmolytic solutions that do not exhibit ion stress on top of
osmotic stress.
Conclusions
Cell wall deposition in tip-growing root hairs is independent
of turgor pressure because it occurs under every
osmotic condition tested. Roots can adapt to osmotic
conditions to a certain extent and can form root hairs also
in hypertonic media by elevation of the osmotic value of
the cells. In future experiments, we hope to gain more

Plasmolysis and cell wall deposition in root hairs 61
insight into the structural needs of cells in order to survive
in extreme environments.
Acknowledgements Partial financial support was given by grant no.
APVV-0432-06 to M.O. from the Grant Agency APVV, Bratislava, SK,
and by grant HJST 1939/2008 of the “Hochschuljubliäumsstiftung der
Stadt Wien” to I.L. Many thanks to Gregor Eder for technical support.
Conflict of interest statement
conflict of interest.
References
The authors declare that they have no
Attree SM, Sheffield E (1985) Plasmolysis of Pteridium protoplasts: a
study using light and scanning-electron microscopy. Planta
165:151–157
Bachewich CL, Heath IB (1997) Differential cytoplasm–plasma
membrane–cell wall adhesion patterns and their relationships to
hyphal tip growth and organelle motility. Protoplasma 200:71–86
Baluška F, Šamaj J, Wojtaszek P, Volkmann D, Menzel D (2003)
Cytoskeleton–plasma membrane–cell wall continuum in plants.
emerging link revisited. Plant Physiol 133:482–491
Barthou H, Petitprez M, Brière C, Souvré A, Alibert G (1999) RGDmediated
membrane–matrix adhesion triggers agarose-induced embryoid
formation in sunflower protoplasts. Protoplasma 206:143–151
Betz WJ, Mao F, Smith CB (1996) Imaging exocytosis and
endocytosis. Curr Opin Neurobiol 6:365–371
Bolte S, Talbot C, Boutte Y, Catrice C, Read ND, Satiat-Jeunemaitre
B (2004) FM-dyes as experimental probes for dissecting vesicle
trafficking in living plant cells. J Microsc 214:159–173
Brett CT (2000) Cellulose microfibrils in plants: biosynthesis, deposition,
and integration into the cell wall. Int Rev Cytol 199:161–199
Buer CS, Weathers PJ, Swartzslander GAJ (2000) Changes in
Hechtian strands in cold-hardened cells measured by optical
mircosurgery. Plant Physiol 122:1365–1377
Canut H, Carrasco A, Galaud J-P, Cassan C, Bouyssou H, Vita N,
Ferrara P, Pont-Lezica R (1998) High affinity RGD-binding sites
at the plasma membrane of Arabidopsis thaliana links the cell
wall. Plant J 16:63–71
de Vries H (1877) Eine Methode zur Analyse der Turgorkraft. Jahrb
wiss Bot 14:427–601
DeBolt S, Gutierrez R, Ehrhardt DW, Somerville C (2007) Nonmotile
cellulose synthase subunits repeatedly accumulate within localized
regions at the plasma membrane in Arabidopsis hypocotyl
cells following 2, 6-dichlorobenzonitrile treatment. Plant Physiol
145:334–338
Denekamp M, Smeekens SC (2003) Integration of wounding and osmotic
stress signals determines the expression of the AtMYB102
transcription factor gene. Plant Physiol 132:1415–1423
Desprez T, Juraniec M, Crowell EF, Jouy H, Pochylova Z, Parcy F,
Höfte H, Gonneau M, Vernhettes S (2007) Organisation of cellulose
synthase complexes involved in primary cell wall synthesis in
Arabidopsis thaliana. Proc Natl Acad Sci U S A 104:15572–15577
Emans N, Zimmermann S, Fischer R (2002) Uptake of a fluorescent
marker in plant cells is sensitive to brefeldin A and wortmannin.
Plant Cell 14:71–86
Galway ME (2000) Root hair ultrastructure and tip growth. In:
Ridge RW, Emons AMC (eds) Root hairs: cell and molecular
biology. Springer, Tokyo, pp 1–17
Galway ME, Heckman JW Jr, Schiefelbein JW (1997) Growth and
ultrastructure of Arabidopsis root hairs: the rhd3 mutation alters
vacuole enlargement and tip growth. Planta 201:209–218
Gouget A, Senchou V, Govers F, Sanson A, Barre A, Rouge P,
Pont-Lezica RP, Canut H (2006) Lectin receptor kinases participate
in protein–protein interactions to mediate plasma membranecell
wall adhesions in Arabidopsis. Plant Physiol 140:81–90
Hara-Nishimura I, Maeshima M (2000) Vacuolar Prozessing Enzymes
and Aquaporins. In: Robinson DR, Rogers JC (eds) Vacuolar
compartments. Sheffield Academic Press, Sheffield, pp 20–42
Hecht K (1912) Studien über den Vorgang der Plasmolyse. Beitr Biol
Pflanz 11:133–189
Hepler PK, Vidali L, Cheung AY (2001) Polarized cell growth in
higher plants. Annu Rev Cell Dev Biol 17:159–187
Hostetter MK (2000) RGD-mediated adhesion in fungal pathogens of
humans, plants and insects. Curr Opin Microbiol 3:344–348
Johansson I, Karlsson M, Johanson U, Larsson C, Kjellbom P (2000)
The role of aquaporins in cellular and whole plant water ballance.
Biochem Biophys Acta 1465:324–342
Johnson-Flanagan AM, Singh J (1986) Membrane deletion during
plasmolysis in hardened and non-hardened plant cells. Plant, Cell
Environ 9:299–306
Jonak C, Kriegerl S, Ligterink W, Barker PJ, Huskisson NS, Hirt H
(1996) Stress signaling in plants: a mitogen-activated protein
kinase pathway is activated by cold and drought. Proc Natl Acad
Sci U S A 93:11274–11279
Kiba A, Sugimoto M, Toyoda K, Ichinose Y, Yamada T, Shiraishi T
(1998) Interaction between cell wall and plasma membrane via
RGD motif is implicated in plant cell defense responses. Plant
Cell Physiol 39:1245–1249
Kohorn BD (2001) WAKs; cell wall associated kinases. Curr Opin
Cell Biol 13:529–533
Lang-Pauluzzi I (2000) The behaviour of the plasma membrane during
plasmolysis: a study by UV microscopy. J Microsc 198:188–198
Lang-Pauluzzi I, Gunning BES (2000) A plasmolytic cycle: the fate of
cytoskeletal elements. Protoplasma 212:174–185
Lang I, Barton DA, Overall RL (2004) Membrane-wall attachements
in plasmolysed plant cells. Protoplasma 224:231–243
Lerch G (1960) Untersuchungen über Wurzelkallose. Botanische
Studien 11:1–111
Lichtscheidl IK, Hepler PK (1996) Endoplasmic reticulum in the
cortex of plant cells. In: Smallwood M, Knox JP, Bowles DJ
(eds) Membranes: specialized functions in plants. BIOS Scientific
Publishers Ltd., Oxford, pp 383–402
Lichtscheidl IK, Url WG (1990) Organization and dynamics of
cortical endoplasmic reticulum in inner epidermal cells of onion
bulb scales. Protoplasma 157:203–215
McNeil M, Darvill AG, Fry SC, Albersheim P (1984) Structure and
function of the primary cell walls in plants. Annu Rev Biochem
53:625–663
Meskiene I, Hirt H (2000) MAP kinase pathways: molecular plugand-play
chips for the cell. Plant Mol Biol 42:791–806
Murai-Hatano M, Kuwagata T (2007) Osmotic water permeability of
plasma and vacuolar membranes in protoplasts I. High osmotic
water permeability in radish (Raphanus sativus) root cells as
measured by a new method. J Plant Res 120:175–189
Oparka KJ (1994) Plasmolysis: new insights into an old process. New
Phytol 126:571–591
Oparka KJ, Prior DAM, Harris N (1990) Osmotic properties of fluidphase
endocytosis in onion epidermal cells. Planta 180: 555–561
Oparka KJ, Wright KM, Murant EA, Allan EJ (1993) Fluid phase
endocytosis: do plants need it? J Exp Bot 44:247–255
Oparka KJ, Prior DAM, Crawford JW (1994) Behaviour of plasma
membrane, cortical ER and plasmoldesmata during plasmolysis
of onion epidermal cells. Plant, Cell Environ 17:163–171
Oparka KJ, Prior DAM, Crawford JW (1996) Membrane conservation
during plasmolysis. In: Smallwood M, Knox JP, Bowles DJ (eds)
Membranes: Specialized Functions in Plants. BIOS Scientific
publishers Ltd, pp 39-56

62 M. Volgger et al.
Ovečka M, Lang I, Baluška F, Ismail A, Illeš P, Lichtscheidl IK (2005)
Endocytosis and vesicle trafficking during tip growth of root
hairs. Protoplasma 226:39–54
Ovečka M, Lichtscheidl IK, Baluška F, Šamaj J, Volkmann D, Hirt H
(2008b) Regulation of root hair tip growth: can mitogen-activated
protein kinases be taken into account? In: Blume YB, Baird WV,
Yemets AI, Breviario D (eds) The plant cytoskeleton: a key tool
for agro-biotechnology. Springer, New York, pp 91–128
Overall RL, Dibbayawan TP, Blackman LM (2001) Intercellular
alignments of the plant cytoskeleton. J Plant Growth Reg 20:
162–169
Paredez AR, Somerville CR, Ehrhardt DW (2006) Visualization of
cellulose synthase demonstrates functional association with
microtubles. Science 312:1491–1495
Parker JS, Cavell AC, Dolan L, Robert K, Grierson CS (2000) Genetic
interactions during root hair morphogenesis in Arabidopsis. Plant
Cell 12:1961–1974
Persson S, Paredez A, Carroll A, Palsdottir H, Doblin M, Poindexter P,
Khitrov N, Auer M, Somerville CR (2007) Genetic evidence for
three unique components in primary cell-wall cellulose synthase
complexes in Arabidopsis. Proc Natl Acad Sci U S A 104:
15566–15571
Pont-Lezica RF, McNally JG, Pickard BG (1993) Wall-to-membrane
linkers in onion epidermis: some hypotheses. Plant, Cell Environ
16:111–123
Ruoslahti E, Pierschbacher MD (1987) New perspectives in cell
adhesion: RGD and integrins. Science 238:491–497
Šamaj J, Ovecka M, Hlavacka A, Lecourieux F, Meskiene I,
Lichtscheidl I, Lenart P, Salaj J, Volkmann D, Bögre L, Baluška
F, Hirt H (2002) Involvement of the mitogen-activated protein
kinase SIMK in regulation of root hair tip growth. EMBO J
21:3296–3306
Šamaj J, Baluška F, Menzel D (2004) New signalling molecules
regulating root hair tip growth. Trends Plant Sci 9:217–220
Schnepf E, Deichgräber G, Bopp M (1986) Growth, cell wall
formation and differentiation in the protonema of the moss,
Funaria hygrometrica: effects of plasmolysis on the developmental
program and its expression. Protoplasma 133:50–65
Schröter K, Sievers A (1971) Wirkung der Turgorreduktion auf den
Golgi-Apparat und die Bildung der Zellwand bei Wurzelhaaren.
Protoplasma 72:203–211
Shi H, Zhu J-K (2002) SOS4, a pyridoxal kinase gene, is required for
root hair development in Arabidopsis. Plant Physiol 129:585–593
Sieberer B, Ketelaar T, Esseling J, Emons A (2005) Microtubules
guide root hair tip growth. New Phytol 167:711–719
Singh J, Iu B, Johnson-Flanagan AM (1987) Membrane alterations in
winter rye and Brassica napus cells during lethal freezing and
plasmolysis. Plant, Cell Environ 10:163–168
Sitte P (1963) Zellfeinbau bei Plasmolyse. Protoplasma 57:304–333
Strobl J, Lichtscheidl IK (2004) Stress performed by heavy metals in
root hairs of Triticum aestivum. In: 14th FESPB Congress. Acta
Physiologiae Plantarum, Special Issue AS-073 Cracow, Poland
Tyerman SD, Niemietz CM, Bramley H (2002) plant aquaporins:
multifunctional water and solute channels with expanding roles.
Plant Cell Env 25(2):173–194
Varner JE, Lin LS (1989) Plant cell wall architechture. Cell 56:231–239
Volkmann D (1984) The plasma membrane of growing root hairs is
composed of zones of local differentiation. Planta 162:392–403
Wang J, Elliott JE, Williamson RE (2008a) Features of the primary
wall CESA complex in wild type and cellulose-deficient mutants
of Arabidopsis thaliana. J Exp Bot 59:2627–2637
Wang Y, Zhang W, Li K, Sun F, Han C, Wang Y, Li X (2008b) Saltinduced
plasticity of root hair development is caused by ion
disequilibrium in Arabidopsis thaliana. J Plant Res 121:87–96

Journal of Experimental Botany, Vol. 57, No. 15, pp. 4201–4213, 2006
doi:10.1093/jxb/erl197 Advance Access publication 3 November, 2006
RESEARCH PAPER
Aluminium toxicity in plants: internalization of aluminium
into cells of the transition zone in Arabidopsis root apices
related to changes in plasma membrane potential,
endosomal behaviour, and nitric oxide production
Peter Illéš 1 , Markus Schlicht 2 ,Ján Pavlovkin 1 , Irene Lichtscheidl 3 , František Baluška 1,2 and
Miroslav Ovečka 1, *
1 Institute of Botany, Slovak Academy of Sciences, Dubravska cesta 14, SK-845 23, Bratislava, Slovakia
2 Institute of Cellular and Molecular Botany, University of Bonn, Bonn, Germany
3 Institution of Cell Imaging and Ultrastructure Research, University of Vienna, Vienna, Austria
Received 6 June 2006; Accepted 11 September 2006
Abstract
The extent of aluminium internalization during the
recovery from aluminium stress in living roots of
Arabidopsis thaliana was studied by non-invasive
in vivo microscopy in real time. Aluminium exposure
caused rapid depolarization of the plasma membrane.
The extent of depolarization depends on the developmental
state of the root cells; it was much more extensive
in cells of the distal than in the proximal portion of
the transition zone. Also full recovery of the membrane
potential after removal of external aluminium was
slower in cells of the distal transition zone than of its
proximal part. Using morin, a vital marker dye for aluminium,
and FM4-64, a marker for endosomal/vacuolar
membranes, an extensive aluminium internalization
was recorded during the recovery phase into endosomal/vacuolar
compartments in the most aluminiumsensitive
cells. Interestingly, aluminium interfered with
FM4-64 internalization and inhibited the formation of
brefeldin A-induced compartments in these cells. By
contrast, there was no detectable uptake of aluminium
into cells of the proximal part of the transition zone and
the whole elongation region. Moreover, cells of the distal
portion of the transition zone emitted large amounts
of nitric oxide (NO) and this was blocked by aluminium
treatment. These data suggest that aluminium internalization
is related to the most sensitive status of the
distal portion of the transition zone towards aluminium.
Aluminium in these root cells has impact on endosomes
and NO production.
Key words: Aluminium internalization, Arabidopsis thaliana,
endosomal compartments, live cell microscopy, membrane
potential, morin vital staining, nitric oxide, recovery, root
transition zone, vacuoles.
Introduction
Aluminium toxicity is an important growth-limiting factor in
acid soils. The main symptom of aluminium toxicity is the
dramatic inhibition of root growth. Some decades ago, two
pioneer works postulated that the decreased root growth is a
consequence of the inhibition of cell division (Clarkson,
1965) and cell elongation (Klimashevski and Dedov, 1975).
Later Ryan et al. (1993) recognized the root apex as a primary
site of aluminium-induced injury in plants. More recently,
numerous reports in the literature describe the aluminiuminduced
changes occurring particularly in the apical regions
of the root, leading to expression of aluminium-toxicity
symptoms: changes in root cell patterning (Doncheva et al.,
2005), irregular cell division, alterations in cell shape, and
vacuolization (Vázquez et al., 1999; Čiamporová, 2000),
* To whom correspondence should be addressed. E-mail: miroslav.ovecka@savba.sk
Abbreviations: BFA, brefeldin A; cPTIO, 2-(4-carboxyphenyl)-4,4,5,5-tetramethylimidazoline-1-oxyl-3-oxide; DAF-2 DA, 4,5-diamino-fluorescein diacetate;
DAG, days after germination; DMSO, dimethylsulphoxide; DTZ, distal part of the transition zone; E D , diffusion potential; E m , plasma membrane potential;
EZ, elongation zone; FM4-64, (N-(3-triethylammoniumpropyl)-4-(8-(4-(diethylamino) phenyl)hexatrienyl)pyridinium dibromide); NO, nitric oxide; PTZ,
proximal part of the transition zone; SHAM, salicylhydroxamic acid.
ª The Author [2006]. Published by Oxford University Press [on behalf of the Society for Experimental Biology]. All rights reserved.
For Permissions, please e-mail: journals.permissions@oxfordjournals.org

4202 Illéš et al.
cell wall thickening and callose deposition (Horst et al., 1999;
Nagy et al., 2004; Jones et al., 2006), disintegration of the cytoskeleton
(Sivaguru et al., 1999, 2003b), formation of myelin
figures and membranous electron-dense deposits (Vázquez,
2002), disturbance of plasma membrane properties (Miyasaka
et al., 1989; Olivetti et al., 1995; Pavlovkin and Mistrík, 1999;
Sivaguru et al., 1999, 2003b; Ahnet al., 2001, 2002, 2004;
Ahn and Matsumoto, 2006), as well as the production of reactive
oxygen species (Darkó et al., 2004; Jones et al., 2006).
These reactions are only few examples of how aluminium
affects the root cells.
The root apex consists of the zone of cell division
(meristem), followed by the distal and proximal transition
zone where cells are prepared for rapid cell expansion
in the elongation zone (Baluška et al., 1990, 1994, 1996;
Ishikawa and Evans, 1993; Verbelen et al., 2006). Sivaguru
and Horst (1998) discovered that the distal part of the transition
zone (DTZ) is the most sensitive part of the root to
aluminium stress. Using a sophisticated experimental approach,
they revealed that local applications of aluminium
to the rapidly elongating cells do not inhibit their growth,
while local applications of aluminium to cells of the distal
portion of the transition zone dramatically inhibit root
growth. However, it remains elusive which processes specific
to this small developmental window in root cell
development are particularly sensitive to aluminium.
Studies by Kollmeier et al. (2000) indicate that the unique
status of auxin in cells of the distal portion of the transition
zone could be responsible for this high sensitivity of DTZ
cells to aluminium.
One of the most relevant problems of recent research on
aluminium phytotoxicity is to define the primary site of its
action at a cellular and subcellular level. The crucial question
is whether aluminium acts primarily in the apoplast or in
the symplast. Consequently, the determination of primary
cellular mechanisms responsible for the rapid cell response
to the aluminium toxicity is still a matter of discussion.
Therefore, it is necessary to characterize the uptake of aluminium
into the root cells and to monitor its spatial and
temporal distribution in cells of living roots.
Vital staining is one of the best and rapid methods for monitoring
aluminium localization and distribution in plants. At
present, despite some negative views (Eticha et al., 2005),
the aluminium-specific fluorescent dye morin as well
as lumogallion seem to be the best vital fluorescent dyes
for aluminium detection. Both of them appear effective in
the nanomolar range of aluminium concentrations. The
morin-staining method showed that aluminium was rapidly
taken up into cultured tobacco BY-2 cells (Vitorello and
Haug, 1996), wheat (Tice et al., 1992), and maize root cells
(Jones et al., 2006). Lumogallion staining confirmed the
rapid aluminium accumulation in soybean root cells (Silva
et al., 2000). On the other hand, Ahn et al. (2002) reported
that most of the aluminium detected by morin was located
preferentially in the cell walls of squash root cells within
the first 3 h of an experiment. Transmission electron microscopy
studies in combination with energy-dispersive X-ray
analysis are approaches giving more detailed information
about the distribution of aluminium at the subcellular
and ultrastructural level. Vázquez et al. (1999) observed
the presence of aluminium in cell walls and vacuoles of
maize root tip cells after 4 h of aluminium treatment.
Interestingly, 24 h of exposure resulted in abundant occurrence
of aluminium deposits within vacuoles, but less in
cell walls. Accelerator mass spectrometry in single cells
of Chara corallina revealed the uptake of aluminium
into the cytoplasm during the first 30 min followed by its
sequestration into vacuoles, although intracellular aluminium
represented only 0.5%; the major portion being
apoplastic (Taylor et al., 2000).
Despite extensive research efforts focusing on aluminium
uptake, results are often conflicting. Most of the discrepancies
arise from different methods and experimental conditions
(concentrations of aluminium, sensitivity of methods,
exposure times, cell types, sample processing, etc). In the
studies on intact roots, the particular stage of cellular development
(Baluška et al., 1996), reflecting its different sensitivity
to aluminium (Sivaguru and Horst, 1998; Sivaguru
et al., 1999), was not always addressed with respect to
aluminium internalization. Therefore, it is difficult to generalize
our knowledge about the uptake of aluminium into
plant cells. Moreover, data on the fate of aluminium internalized
during plant recovery are completely missing.
To gain a synoptic understanding of aluminium effects,
the extent of aluminium internalization into well-defined
cells of living roots of Arabidopsis thaliana was studied
in a continuous mode under controlled conditions by noninvasive
live microscopy. It is reported that those cells
which are most aluminium-sensitive are also the most
active in the internalization of apoplastic aluminium
during recovery. Importantly, endosomal and vacuolar compartments
are highly enriched with the internalized
aluminium only in cells of the distal portion of the transition
zone.
Materials and methods
Plant material and growth conditions
Seeds of Arabidopsis thaliana L., ecotype Columbia, were surfacesterilized
with 0.25% sodium hypochlorite for 3 min, washed and sown
on an agar-solidified nutrient medium in Petri dishes. The nutrient medium
was based on Murashige-Skoog salts (Murashige and Skoog,
1962) with addition of vitamins (myo-inositol 10 mg l
1 , calcium
pantothenate 0.1 mg l
1 , niacin 0.1 mg l
1 , pyridoxin 0.1 mg l
1 ,
thiamin 0.1 mg l
1 , biotin 0.001 mg l
1 of medium), FeSO 4 .7H 2 O
(1.115 mg l
1 ), CaCl 2 (111 mg l
1 ), sucrose (10 g l
1 ), and agar
(10gl
1 ), the final pH was adjusted to 4.5. The seeds were vernalized
at 4 °C for 24 h. Petri dishes were placed into a growth chamber,
positioned vertically and kept under controlled environmental
conditions at 25 °C, 180 lmol m 2 s 1 and a 12/12 h day/night
rhythm.

Aluminium sensitivity of root cells related to aluminium internalization 4203
Root growth and detection of aluminium
Effects of aluminium on root growth were observed on seedlings
which, 2 d after germination (DAG), were transferred to Petri dishes
containing agar-solidified nutrient medium with different aluminium
concentrations (0, 10, 50, 100, 200, and 300 lM total concentration
of Al in the form of AlCl 3 .6H 2 O). Root elongation was measured
every day during a 7-d period. After 7 d of cultivation, aluminium
was detected by staining whole roots with haematoxylin (Polle
et al., 1978) or morin (Vitorello and Haug, 1997). Roots stained with
haematoxylin were observed in bright field and morin fluorescence
was visualized by an Olympus BX51 microscope (Olympus, Japan)
equipped with BP 470–490 excitation filter, BA 515 IF barrier filter
and a DM 505 dichroic mirror.
Electrophysiology
Measurements of the plasma membrane electrical potential difference
(E m ) were carried out at 24 °C in root cells of intact Arabidopsis seedlings,
2 DAG, by the standard microelectrode technique as described
by Pavlovkin et al. (1993). The perfusion solution contained 0.1 mM
KCl, 1 mM Ca(NO 3 ) 2 , 1 mM NaH 2 PO 4 , 0.5 mM MgSO 4 ; pH was
adjusted to 4.5. Diffusion potential (E D ) was determined by application
of inhibitors (1 mM NaCN+1 mM SHAM) dissolved in perfusion
solution. The influence of morin on E m was measured upon
adding 100 lM morin to the perfusion solution. Effects of aluminium
on E m were monitored during continual exchange of the perfusion
solution by the experimental solution (50 lM AlCl 3 ) in perfusion
solution, perfusion speed 5 ml min 1 ). After 5 min or 30 min of aluminium
exposure the experimental solution was washed out with
perfusion solution. Changes of E m induced by aluminium were
measured continuously during the whole experiment. Insertion of
the microelectrode into the cortical cells of the distal portion of the
transition zone, located 150–300 lm behind the root tip, and of the
proximal portion of the transition zone, located 300–400 lm behind
the root tip (Verbelen et al., 2006), was performed under microscopic
control. Measurements were evaluated separately for each zone.
Live microscopy of aluminium internalization into the cells
Arabidopsis seedlings 2 DAG were transferred to microscopic slides
modified to micro-chambers by coverslips (Ovečka et al., 2005).
The chambers were filled with liquid nutrient medium of the same
composition as used for cultivation in Petri dishes, but without agar
and placed into sterile glass cuvettes containing the same nutrient
medium (pH 4.5). The seedlings were grown in a vertical position
under light in the growth chamber. During a 12-h period the seedlings
resumed stable root growth. Subsequently the micro-chambers
were gently perfused with the aluminium-containing medium
(50 lM AlCl 3 in nutrient solution, pH 4.5, perfusion speed 10 ll
min 1 ). After a 30 min pulse treatment, aluminium was washed out
and the roots of the Arabidopsis seedlings were stained with 100
lM morin for 20 min, using the same perfusion technique and then
washed with the nutrient solution. After this exchange, internalization
experiments were performed in the micro-chambers directly on the
microscope stage during 3 h. For labelling endosomes and tonoplast,
the styryl dye FM4-64 (N-(3-triethylammoniumpropyl)-4-(8-(4-
(diethylamino) phenyl)hexatrienyl)pyridinium dibromide) was used
at a final concentration of 4 lM in nutrient solution, applied for 5
min prior to the aluminium treatment. Internalization of aluminium
wasobservedinlivingrootsinrealtimebyaninvertedmicroscopeLeica
DMIRE2,equippedwith theconfocal laser scanningsystemLeica TCS
SP2 (Leica Microsystems Heidelberg, Germany). Excitation wavelength
for FM4-64 was 514 nm and for morin 458 nm. Fluorescence
was observed between 640 nm and 700 nm (FM4-64) and 480 nm
and 510 nm (morin).
FM4-64 internalization and BFA-induced compartment
formation
Working solution of 5 lM FM4-64 was prepared from the stock solution
(1 mg ml
1 FM4-64 in DMSO). With FM4-64 the plants were
incubated for 10 min and the dye was washed out before observation.
Roots labelled for 5 min with the FM 4-64 were pretreated with
BFA applied at a concentration of 35 lM for 25 min. The effect of
aluminium was studied by the application of 90 lM AlCl 3 for
90 min before treatment with BFA and FM4-64.
NO labelling and measurements in control and
aluminium-treated root cells
Detection of nitric oxide (NO) was performed by the specific fluorescent
probe 4,5-diaminofluorescein diacetate (DAF-2 DA; Calbiochem,
USA). The roots were incubated with 15 lM DAF-2 DA for
30 min and washed before observation. As a negative control, they
were treated with 10 lM of the NO-scavenger cPTIO (Lombardo
et al., 2006). Examination was performed with a confocal laser
scanning microscope system using standard filters and collection
modalities for DAF-2 green fluorescence (excitation 495 nm; emission
515 nm). Fluorescence intensity was measured with the open
source software Image-J (http://rsb.info.nih.gov/ij/).
Results
Influence of aluminium on root elongation and
morphology
Arabidopsis seedlings cultivated for 7 d on agar plates with
different concentrations of AlCl 3 exhibited concentrationdependent
inhibition of root growth (Fig. 1). Growth of
the primary roots was only slightly reduced by 10 lM AlCl 3
(to 95% of the control values, not statistically significant
according to t test at P¼0.05). Inhibition of root growth
was statistically significant at 50 lM AlCl 3 as compared to
control according to t test at P¼0.01. However, growth of
the primary roots was reduced only to 89% of the control.
The inhibition was more apparent at 100 lM and 200 lM
AlCl 3 (59% and 45% of control growth, respectively, highly
significant at P¼0.001), while root growth was fully inhibited
Fig. 1. Root growth of Arabidopsis plants cultivated for 7 d on agar
plates with different concentrations of AlCl 3 . Growth of the primary roots
is progressively affected by 100 lM and 200 lM AlCl 3 while root
elongation is inhibited completely at 300 lM AlCl 3 . Average of 20
seedlings per treatment (6SD).

4204 Illéš et al.
Fig. 2. Effect of aluminium on elongation and morphology of the root system of Arabidopsis seedlings after 4 d cultivation. Position of root tips after
transfer of seedlings to aluminium-containing agar plates is indicated by arrows. Inhibition of root elongation at 200 lM and 300 lM AlCl 3 is
accompanied by enhanced formation of lateral roots. Representative of 20 seedlings per treatment.
Fig. 3. Histochemical detection of aluminium in the roots of Arabidopsis after 7 d cultivation on agar plates by two different staining methods:
haematoxylin staining (A) and morin staining (B). Haematoxylin stained aluminium strongly at 100 lM and higher concentrations of AlCl 3 while the
reaction of morin to aluminium started to be strong at 50 lM AlCl 3 . Note the radial expansion of root cells at 100 lM and 200 lM AlCl 3 , but not at
300 lM AlCl 3 due to the complete inhibition of root growth. Representative of five seedlings per treatment. Bar¼100 lm.
by 300 lM AlCl 3 (only 2% root elongation as compared
to control plants, highly significant at P¼0.001; Figs 1, 2).
Severe inhibition of root elongation was accompanied by radial
expansion of the root cells at 100 lM and 200 lMAlCl 3 ,
but not at 300 lM AlCl 3 , due to the complete inhibition of root
growth (Fig. 3). Root architecture of Arabidopsis seedlings
exposed to lower concentrations of aluminium (10–100
lM AlCl 3 ) was almost unaffected. However, 200 lM and

300 lM AlCl 3 induced enhanced formation of lateral roots
(Fig. 2). Thus, the inhibition of root elongation was tightly
correlated with, and partly compensated by, an increasing
number of growing zones with the formation of lateral roots.
Histochemical detection of aluminium
For histochemical detection of aluminium in the roots of
Arabidopsis plants cultivated on agar plates with different
concentrations of aluminium, two different staining methods
were used: haematoxylin staining and morin staining (Polle
et al., 1978; Vitorello and Haug, 1997). Using both methods,
an aluminium-specific signal was observed mainly in the
apical parts of the roots exposed to aluminium for 7 d
(Fig. 3). In the control (Fig. 3), and at 10 lM AlCl 3 (data
not shown), aluminium staining by haematoxylin was
not detectable and only a weak diffuse fluorescence of morin
occurred. At 50 lM AlCl 3 , there was hardly any detection
of aluminium in roots by means of haematoxylin, while
morin gave bright fluorescence (Fig. 3). At higher concentrations,
haematoxylin started to detect aluminium in the
roots, the pattern of staining being similar to the morin fluorescence.
In the roots grown at 100 lM and 200 lM
AlCl 3 , both staining methods showed maximum accumulation
of aluminium in the root apex. In the plants exposed
to 300 lM AlCl 3, aluminium was abundant not
only in the root apex but it also invaded the root central
cylinder (Fig. 3).
Obviously, the sensitivity of morin to aluminium is higher
than that of haematoxylin. Thus, morin is considered as
a more suitable tracer dye for aluminium detection in the
cells of the Arabidopsis root apex grown on aluminiumsupplemented
agar plates. The critical discrimination limit
in the sensitivity between morin and haematoxylin is apparently
at 50 lM AlCl 3 (Fig. 3). Because of mild effects on
root growth and the capability of morin to localize aluminium,
50 lM AlCl 3 was utilized as the indicative testing
concentration for monitoring aluminium effects on
Arabidopsis roots in the following experiments.
Aluminium sensitivity of root cells related to aluminium internalization 4205
on this electrophysiological characteristic, all further experiments
were performed separately for DTZ and PTZ.
The diffusion potential (E D ) was determined in order
to distinguish between passive and active, i.e. energydependent,
components of the E m by application of inhibitors
(1 mM NaCN+1 mM SHAM); the E m of cortical cells
rapidly depolarized in both DTZ and PTZ to E D ( 40 to 41
mV, Fig. 4). As a control, 100 lM morin revealed almost
no detectable changes of E m (data not shown); thus, morin
does not significantly affect the plasma membrane properties.
This confirms the suitability of this dye for studying
aluminium distribution in living cells.
The main objective of the present electrophysiological
experiments was to characterize dynamic changes of E m
and to determine the sensitivity of the two developmental
zones (DTZ and PTZ) during the exposure to 50 lM AlCl 3 .
Aluminium-induced depolarization of E m occurred within
2 min after aluminium application in both developmental
zones (Fig. 5A). However, the membrane potential depolarized
further in the cells of DTZ (to E D ) than in the cells of
PTZ ( 78 mV to 88 mV). Complete repolarization of E m
was achieved by removing aluminium from the perfusion
solution within 10 min in the cells of DTZ, while the cells
of PTZ repolarized within only 3 min (Fig. 5A). While the
period of treatment used in the aluminium internalization
experiments was 30 min, the effects of short-term aluminium
treatment (5 min; Fig. 5A) were compared with the
30 min exposure (Fig. 5B). Extent and speed of the E m depolarization
were similar (data not shown), but the speed
of repolarization was again different in the two developmental
zones. After aluminium removal from the solution,
the time required for complete repolarization was 14 min in
the cells of DTZ and only 6 min in the cells of PTZ (Fig. 5B).
Apparently DTZ is much more sensitive to aluminium than
PTZ. Nevertheless, depolarization of the plasma membrane
Effects of aluminium on the plasma membrane
electrical potential
In order to detect immediate cell responses of the root apex
to aluminium, the plasma membrane potential (E m ) was
recorded before and during aluminium application, as well
as after the removal of aluminium by washing. E m in the cells
of the root cortex revealed distinct properties in different
developmental zones. In this case, the distal part of the transition
zone, just behind the cell division zone, was located
150–300 lm behind the root tip, the proximal part of the
transition zone 300–400 lm behind the root tip (Verbelen
et al., 2006). E m of cortical cells varied between 82 mV
and 98 mV ( 8864 mV, n¼37) in the distal part of the
transition zone (DTZ) and between 94 mV and 117 mV
( 10567 mV, n¼29) in the proximal part (PTZ). Based
Fig. 4. Effect of 1 mM NaCN and 1 mM salicylhydroxamic acid
(SHAM) on cortical cell membrane potential (E m ). Both in proximal
transition zone (A) and distal transition zone (B), the E m rapidly
depolarized to the values of diffusion potential (E D ). Representative of
20 seedlings per treatment.

4206 Illéš et al.
Fig. 5. Changes of cortical cell membrane potential (E m ) after treatment
with 50 lM AlCl 3 . E m depolarized rapidly (within 2 min) after aluminium
application; depolarization was more extensive in the cells of distal
transition zone (DTZ) than in the proximal transition zone (PTZ). After
removing aluminium, E m in the PTZ completely repolarized within 3 min,
while in the DTZ within 10 min (A). After 30 min aluminium treatment,
the complete repolarization of E m occurred within 6 min in the cells of
PTZ and within 14 min in the DTZ (B). Representative of 24 seedlings
per treatment.
was fully reversible, which was consequently followed by
recovery of root growth (data not shown).
Internalization and accumulation of aluminium within
endosomal/vacuolar compartments
The dynamics of aluminium internalization was monitored
in the root cells of Arabidopsis seedlings 2 d after germination.
The intact roots were treated with 50 lM AlCl 3 for
30 min, washed and stained with morin. After washing out
morin, the seedlings were kept in control medium for recovering
and the time-course of aluminium internalization in the
living root apices could be observed with a confocal microscope
for 3 h. In cells of meristem and DTZ, aluminium was
located exclusively in the apoplast during the first 20 min
of recovery (after aluminium removal; Fig. 6A). From then
on, it was internalized into the cells. A first detectable diffuse
fluorescence signal of morin-stained aluminium in the cytoplasm
appeared after 60 min (Fig. 6B). Aluminium further
proceeded into roundish structures with blurred edges, and
2 h and 30 min after the end of treatment it was located in
vacuole-like structures of different size with clearly defined
boundaries. In the cytoplasm the signal became weaker, in
the cell walls it remained present (Fig. 6C). After 3 h and 30
min, these cells accumulated aluminium almost exclusively
in the vacuole-like compartments (Fig. 6D).
The assumption that the target compartment of aluminium
sequestration in the cells is the vacuole was confirmed
by FM4-64, the dye widely used for labelling the plasma
membrane, endocytic membranous compartments and
tonoplast (Betz et al., 1996; Geldner, 2004; Ovečka et al.,
2005; Šamaj et al., 2005). Accumulation of aluminium
was shown in the cells of root developmental zones that,
as evident from the previous experiments, revealed different
sensitivity to aluminium (Fig. 5). In the meristematic cells
(Fig. 7A) and in the cells of the distal portion of the transition
zone (Fig. 7B), aluminium was internalized rapidly
and accumulated into the vacuolar compartments within
2 h and 50 min of recovery. Tonoplast was labelled red with
FM4-64 whereas the aluminium-containing lumen of the
vacuoles was stained green with morin (Fig. 7A, B).
In the cells of the proximal portion of the transition zone,
there was no prominent accumulation of aluminium even
3 h 10 min after aluminium deprivation (Fig. 7C). Moreover,
aluminium did not enter vacuoles neither in the proximal
portion of the transition zone nor in the elongation
zone, it could only be detected in the apoplast. This indicates
that after the removal of free aluminium ions from the medium,
residual aluminium bound to the cell wall can be internalized
into endosomal compartments and vacuoles of
living cells of the meristem as well as of the distal portion
of the transition zone. On the contrary, in the less sensitive
cells of the proximal portion of the transition zone, as well as
in the elongation region (data not shown), cell wall-bound
aluminium is not internalized and remains located in the
apoplast.
Effects of aluminium on endosomal behaviour
After 10 min exposure of roots, FM4-64 strongly labelled
cross-walls of the cells as well as the early endosomes
(Fig. 8A; Voigt et al., 2005). When such cells were exposed
to brefeldin A (BFA), the early endosomes aggregated
into the so-called BFA-induced compartments (Fig. 8B;
Šamaj et al., 2004, 2005). Aluminium treatment prevented
formation of such BFA-induced compartments (Fig. 8C, D),
suggesting that early endosomes, involved in aluminium
internalization, were affected.
Effects of aluminium on NO production
NO-specific DAF-2DA labelling showed spatial distribution
of NO production in cells of control root tips (Fig. 9A)
and local changes of this distribution induced by aluminium

Aluminium sensitivity of root cells related to aluminium internalization 4207
Fig. 6. Time-course of aluminium uptake in the cells of the meristem and DTZ monitored by morin. Pulse treatment of Arabidopsis roots with 50 lM
AlCl 3 for 30 min, followed by washing and morin staining. Observation of the root cells 20 min after the end of aluminium treatment revealed the
presence of aluminium only in the apoplast (A). The first signals of morin fluorescence in the cytoplasm were detected within 1 h (B). 2 h and 30 min after
treatment aluminium accumulated into roundish vacuole-like structures of varying size (C). 3 h 30 min after treatment aluminium was sequestered in
vacuole-like compartments (D). Representative of five seedlings per treatment. Bar¼10 lm.
treatment (Fig. 9B). After application of the NO scavenger
cPTIO, the DAF-2DA fluorescence signal was lacking, indicating
effective NO scavenging (Fig. 9C). In control root
apices, there were three local centres of NO production:
one at the root cap statocytes, another one at the quiescent
centre and distal portion of the meristem, and the third, the
most prominent one, at the distal part of the transition zone
(the blue line in Fig. 9D). While the NO-scavenger stopped
NO production in roots (the green line in Fig. 9D), aluminium
treatment (90 lM for 1 h) completely abolished only the NO
production peak in the distal part of the transition zone; the
first two peaks became even more pronounced (the red line
in Fig. 9D).
Discussion
Although aluminium toxicity in plants has been extensively
studied from different points of view, a complete image
of its distribution at the cellular level is still missing. In line
with the supporting data for aluminium uptake into the
cells, evidence for predominant accumulation of aluminium
only in the apoplast has also been given. However, it must be
kept in mind that much of the data available in this field were
obtained with different plant species and under experimental
conditions which were not comparable. Experimental
approaches such as the detection of aluminium in living cells
with high sensitivity in physiological conditions may contribute
to clarifying this situation.
Internalization and distribution of aluminium can be
visualized at low concentrations in living cells
The time-course of internalization of aluminium in actively
growing root cells of Arabidopsis thaliana was detected by
the application of non-invasive microscopy techniques.
Both dyes morin and FM4-64 were used as vital markers
in living cells under microscopic control. While FM4-64
is widely used as a vital marker of endocytosis in different
cell types (Vida and Emr, 1995; Betz et al., 1996; Fischer-
Parton et al., 2000; Bolte et al., 2004; Ovečka et al., 2005;
Šamaj et al., 2005; Dettmer et al., 2006; Dhonukshe et al.,
2006), morin was mainly used as a last staining step in the
final stage of the experiments. In this study, morin was
used as a marker of aluminium redistribution in living Arabidopsis
root cells. Morin labelling in the early stages of

4208 Illéš et al.
Fig. 7. Internalization of aluminium in specific developmental zones of pulse-treated Arabidopsis roots with 50 lM AlCl 3 for 30 min. Roots were
pretreated with 4 lM FM4-64 and stained with morin after washing out the aluminium. In the meristematic cells (A) and the cells of distal transition zone
(B) aluminium was internalized and accumulated in vacuolar compartments after 2 h 50 min of recovery. Tonoplast was labelled red with FM4-64 and
the aluminium-containing lumen of the vacuole was stained green with morin. In the proximal transition zone (C) there was no uptake of aluminium even
3 h 10 min after the end of the treatment; aluminium was not accumulated in the vacuoles and could be detected only in the apoplast. Representative of
five seedlings per treatment. Bar¼10 lm.
recovery and careful visualization of its fluorescence
allowed the time-course of aluminium internalization to
be studied at low, non-lethal concentrations even in the
most sensitive cells of DTZ.
Cells of various root developmental zones have
different sensitivity to aluminium and show
specific patterns of recovery
The effect of aluminium on root cells became evident by
rapid changes of the electrical membrane potential, which
were different in the cells of various developmental stages.
The root apex consists of distinct developmental zones
including the cell division zone (meristem), two zones of
preparation for rapid cell expansion (DTZ and PTZ), followed
by the actual zone of rapid cell elongation (Baluška
et al., 1990, 1994, 1996; Ishikawa and Evans, 1993; Verbelen
et al., 2006). Aluminium caused the rapid depolarization of
the plasma membrane electro-potential (E m ) in the cells of
both the DTZ and PTZ. The extent of depolarization, however,
was much greater in the more sensitive DTZ. This is
in accordance with the observations by Sivaguru and Horst
(1998), Horst et al. (1999), and Sivaguru et al. (1999,
2003a), who described a different sensitivity of the cells
in different developmental zones. This clearly indicates that
the extent of the aluminium sensitivity as a function of cellular
developmental stages should be taken into consideration.
Concerning recovery from aluminium stress, removing
free aluminium from the medium was followed by full
regeneration of the E m values. Hence, the changes in
electrophysiological properties of the plasma membrane

Aluminium sensitivity of root cells related to aluminium internalization 4209
Fig. 8. Effect of aluminium on internalization of endocytic marker FM4-64. Control root after 10 min labelling with FM4-64 dye (A). Formation of
BFA-induced compartments by 35 lM BFA in FM4-64-labelled roots (B). Treatment with 90 lM aluminium for 90 min did not change considerably the
pattern of FM4-64 labelling (C), but it prevented formation of BFA-induced compartments after application of 35 lM BFA (D). Representative of five
seedlings per treatment. Bar¼10 lm.
induced by aluminium were reversible under the experimental
conditions in the recovering cells of both DTZ and PTZ.
Consistent with developmentally dependent differences in
sensing aluminium, the process of plasma membrane recovery
was slower in the cells of the DTZ as compared to those
of PTZ.
The cellular distribution of aluminium
Aluminium either accumulates on the cell surface in the
cell walls (Horst et al., 1999; Marienfeld et al., 2000; Wang
et al., 2004) or it enters the cells (Tice et al., 1992; Lazof
et al., 1994, 1996; Vázquez et al., 1999; Silva et al.,
2000; Jones et al., 2006) during exposure to aluminium.
However, information about the fate of aluminium associated
with cell surfaces during recovery is missing. It is shown
here that root cells can restore membrane functions in recovery
experiments. The restoration of membrane functions
together with the removal of the critical aluminium from
the cell surface via its internalization and sequestering within
the vacuole may contribute to the recovery of the growth.
This scenario has been proposed in the present study. The
vacuolar deposits in aluminium-treated maize roots support
the tentative conclusion that vacuolar localization of
the internalized aluminium might be the mechanism of its
intracellular detoxification (Vázquez et al., 1999).
Interestingly, the high rate of aluminium internalization
was typical only for meristematic cells and for the cells of
the distal portion, but not of the proximal portion of the
transition zone. Extracellular aluminium is mainly associated
with cell wall pectins as was manifested by the
correlation between the pectin content in the cell walls and
the accumulation of aluminium (Horst et al., 1999; Schmohl
and Horst, 2000; Hossain et al., 2006). It is speculated at
this early stage that the internalization of aluminium into
the cells might be closely related to the endocytosis of cell
wall pectins. However, consistent with the pattern of aluminium
internalization, internalization and recycling of cell wall
pectins is also accomplished only in the cells of the meristem
and the distal portion of the transition zone, but not
in the region of rapid cell elongation (Baluška et al., 2002,
2005a; Yuet al., 2002; Paciorek et al., 2005; Dhonukshe
et al., 2006). Indeed, endocytosis was active during the recovery
phase as proved by the internalization of the endocytic
marker FM4-64. Endocytosis proceeded in all cells

4210 Illéš et al.
Fig. 9. Detection of NO production by DAF-2DA labelling in control root tip (A), root tips treated with 90 lM aluminium for 60 min (B), and 10 lM
cPTIO, the NO-scavenger, for 60 min (C). Fluorescence of DAF-2DA is green, FM4-64 is red. Fluorescence intensity (D) and distribution (insert) of
DAF-2DA labelling along root developmental zones. Note the disappearance of the NO production peak in DTZ after aluminium treatment. Average
intensities of 20 roots per treatment.
of the root apex including the PTZ and the elongation zone
(see FM4-64 labelling of the tonoplast), although internalization
of aluminium was spatially restricted to the pectinrecycling
zone (Baluška et al., 2002), and did not occur
in the PTZ and the elongation zone. Inside the cells,
endosomal-sorting processes might be implicated in releasing
the aluminium from the pectin complexes; pectins would
be recycled back to the cell wall while aluminium would
continue in the endocytic pathway towards the vacuole. Pectin
molecules reaching the cell wall may represent a new
pool for binding of free and loosely bound apoplastic aluminium
and thus support the gradual removal of morinstained
apoplastic aluminium during recovery. Supporting
data for such endocytic internalization of aluminium come
from the presence of aluminium in myelin figures (Vázquez,
2002), which closely resemble the multilamellar endosomes
occurring in the pectin internalization pathway (Baluška
et al., 2005a). Difficulties with the visualization of these intermediary
structures under experimental conditions could
be caused by weakening of the morin fluorescent signal
intensity. A decrease of the fluorescent signal in the morin
detection method after aluminium binding to pectins
in vitro was shown by Eticha et al. (2005). However, this
finding obtained by both the use of hand sections and

fluorometric analyses of Al sorption to derived cell walls
does not necessarily need to reflect the situation in intact
roots of Arabidopsis exposed to morin.
Impact of aluminium on NO production and polar
transport processes
The data reveal that internalization of aluminium into plant
endosomes alters their behaviour as they fail to form the
BFA-induced compartments. Moreover, this endosomal
aluminium might also influence nitric oxide (NO) production,
which showed its maximum in the cells of DTZ in
control root apices but was suppressed after aluminium
treatment. In animal cells, NO is active in endosomes involved
in the processing of internalized heparan sulphate
(Cheng et al., 2002). In addition, NO regulates endocytosis
and vesicle recycling especially at neuronal synapses
(Meffert et al., 1996; Huang et al., 2005; Kakegawa and
Yuzaki, 2005; Wang et al., 2006). In this respect it is most
interesting that plant synapses (Baluškaet al., 2005b), which
are very active in both endocytosis and vesicle recycling
(Baluška et al., 2003, 2005b), are located exactly in the
aluminium-sensitive distal portion of the transition zone
(Sivaguru and Horst, 1998; Sivaguru et al., 1999).
Finally, there are obvious links between the aluminium
toxicity and the inhibition of the basipetal polar transport
of auxin in the epidermis and outer cortex cells. Hasenstein
and Evans (1988) were the first to discover that aluminium
inhibits the basipetal flow of auxin. Later, this result was
fully confirmed by Kollmeier et al. (2000) who also showed
that the distal portion of the transition zone is the most relevant
in the aluminium-based inhibition of basipetal auxin
transport. Interestingly, the cells in this zone are unique with
respect to auxin and its role in cell growth regulation
(Ishikawa and Evans, 1993; Baluška et al., 1994, 1996,
2004). Doncheva et al. (2005) documented that, in an aluminium-sensitive
maize line, aluminium treatment mimics
auxin transport inhibitors in their morphogenic effects on
cell division planes in the PTZ (see also Dhonukshe
et al., 2005). Polar auxin transport is insensitive to aluminium
in aluminium-tolerant mutant AlRes4 of tobacco (Ahad
and Nick, 2006).
The DTZ cells are not only the most sensitive towards
aluminium toxicity (Sivaguru and Horst, 1998; Sivaguru
et al., 1999), but are also the most active ones in the
cell-to-cell transport of auxin (Mancuso et al., 2005;
Santelia et al., 2005). Recently published data revealed
that polar auxin transport is linked to active vesicle trafficking
and that auxin is secreted out of cells via vesicle recycling
(Schlicht et al., 2006). It is shown here that internalized
aluminium affects the behaviour of endosomes as well as
the production of NO. This indicates that the extraordinary
sensitivity of the DTZ cells towards aluminium could be a
consequence of an extremely active vesicle recycling driving
extensive polar auxin transport in this particular root
apex zone.
Aluminium sensitivity of root cells related to aluminium internalization 4211
Acknowledgements
The authors thank Milada Čiamporova for critical comments on the
manuscript. This work was supported in part by the Grant Agency
VEGA (Grants nos 2/5085/25, 2/5086/25, and 2/3051/23). MO was
supported by a Marie Curie European Reintegration Grant No.
MERG-CT-2005–031168 within the 6th European Community
Framework Programme. Financial support by grants from the
Deutsches Zentrum für Luft- und Raumfahrt (DLR, Cologne,
Germany; project 50WB 0434), from the European Space Agency
(ESA-ESTEC Noordwijk, The Netherlands; MAP project AO-
99–098), and from the Ente Cassa di Risparmio di Firenze (Italy) is
gratefully acknowledged too.
References
Ahad A, Nick P. 2006. Actin is bundled in activation-tagged tobacco
mutants that tolerate aluminium. Planta DOI 10.1007/s00425–
006–0359–0.
Ahn SJ, Matsumoto H. 2006. The role of the plasma membrane in
the response of plant roots to aluminum toxicity. Plant Signaling
and Behavior 1, 37–45.
Ahn SJ, Rengel Z, Matsumoto H. 2004. Aluminium-induced
plasma membrane surface potential and H + -ATPase activity in
near-isogenic wheat lines differing in tolerance to aluminum.
New Phytologist 162, 71–79.
Ahn SJ, Sivaguru M, Chung GC, Rengel Z, Matsumoto H.
2002. Aluminium-induced growth inhibition is associated with
impaired efflux and influx of H + across the plasma membrane in
root apices of squash (Cucurbita pepo). Journal of Experimental
Botany 53, 1959–1966.
Ahn SJ, Sivaguru M, Osawa H, Chung GC, Matsumoto H.
2001. Aluminum inhibits the H + -ATPase activity by permanently
altering the plasma membrane surface potentials in squash roots.
Plant Physiology 126, 1381–1390.
Baluška F, Barlow PW, Kubica Š. 1994. Importance of the postmitotic
‘isodiametric’ growth (PIG) region for growth and development
of roots. Plant and Soil 167, 31–42.
Baluška F, Hlavačka A, Šamaj J, Palme K, Robinson DG,
Matoh T, McCurdy DW, Menzel D, Volkmann D. 2002.
F-actin dependent endocytosis of cell wall pectins in meristematic
root cells. Insights from brefeldin A-induced compartments. Plant
Physiology 130, 422–431.
Baluška F, Kubica Š, Hauskrecht M. 1990. Postmitotic ‘isodiametric’
cell growth in the maize root apex. Planta 181,
269–274.
Baluška F, Liners A, Hlavačka A, Schlicht M, Van Cutsem P,
McCurdy DW, Menzel D. 2005a. Cell wall pectins and xyloglucans
are internalized into dividing root cells and accumulate
within cell plates during cytokinesis. Protoplasma 225, 141–155.
Baluška F, Mancuso S, Volkmann D, Barlow PW. 2004. Root
apices as plant command centres: the unique ‘brain-like’ status of
the root apex transition zone. Biologia 59, 9–17.
Baluška F, Šamaj J, Menzel D. 2003. Polar transport of auxin:
carrier-mediated flux across the plasma membrane or neurotransmitter-like
secretion? Trends in Cell Biology 13, 282–285.
Baluška F, Volkmann D, Barlow PW. 1996. Specialized zones
of development in roots: view from the cellular level. Plant
Physiology 112, 3–4.
Baluška F, Volkmann D, Menzel D. 2005b. Plant synapses: actinbased
adhesion domains for cell-to-cell communication. Trends in
Plant Sciences 10, 106–111.
Betz WJ, Mao F, Smith CB. 1996. Imaging exocytosis and
endocytosis. Current Opinion in Neurobiology 6, 365–371.

4212 Illéš et al.
Bolte S, Talbot C, Boutte Y, Catrice C, Read ND, Satiat-
Jeunemaitre B. 2004. FM-dyes as experimental probes for
dissecting vesicle trafficking in living plant cells. Jornal of
Microscopy 214, 159–173.
Cheng F, Mani K, Van den Born J, Ding K, Belting M,
Fransson L-A. 2002. Nitric oxide-dependent processing of
heparan sulfate in recycling S-nitrosylated glypican-1 takes place
in caveolin-1-containing endosomes. Journal of Biological
Chemistry 277, 44431–44439.
Čiamporová M. 2000. Diverse response of root cell structure to
aluminium stress. Plant and Soil 226, 113–116.
Clarkson DT. 1965. The effect of aluminium and some other
trivalent metal cations on cell division in the root of Allium cepa.
Annals of Botany 29, 309–315.
Darkó É, Ambrus H, Stefanovits-Bányai É, Fodor J, Bakos F,
BarnabásB.2004. Aluminium toxicity, Al tolerance and oxidative
stress in an Al-sensitive wheat genotype and in Al-tolerant lines
developed by in vitro microspore selection. Plant Science 166,
583–591.
Dettmer J, Hong-Hermesdorf A, Stierhof Y-D, Schumacher K.
2006. Vacuolar H + -ATPase activity is required for endocytic and
secretory trafficking in Arabidopsis. The Plant Cell 18, 715–730.
Dhonukshe P, Baluška F, Schlicht M, Hlavačka A, Šamaj J,
Friml J, Gadella Jr TWJ. 2006. Endocytosis of cell wall surface
material mediates cell plate formation during plant cytokinesis.
Developmental Cell 10, 137–150.
Dhonukshe P, Mathur J, Hülskamp M, Gadella Jr TWJ. 2005.
Microtubule plus-ends reveal essential links between intracellular
polarization and localized modulation of endocytosis during
division-plane establishment in plant cells. BMC Biology 3, 11.
Doncheva S, Amenós M, Poschenrieder C, Barceló J. 2005.
Root cell patterning: a primary target for aluminium toxicity in
maize. Journal of Experimental Botany 56, 1213–1220.
Eticha D, Stass A, Horst WJ. 2005. Localization of aluminium in
the maize root apex: can morin detect cell wall-bound aluminium?
Journal of Experimental Botany 56, 1351–1357.
Fischer-Parton S, Parton RM, Hickey P, Dijksterhuis J,
Atkinson HA, Read ND. 2000. Confocal microscopy of FM4-
64 as a tool for analysing endocytosis and vesicle trafficking in
living fungal hyphae. Journal of Microscopy 198, 246–259.
Geldner N. 2004. The plant endosomal system: its structure and role in
signal transduction and plant development. Planta 209, 547–560.
Hasenstein KH, Evans ML. 1988. Effect of cations on hormone
transport in primary roots of Zea mays. Plant Physiology 86,
890–894.
Horst WJ, Schmohl N, Kollmeier M, Baluška F, Sivaguru M.
1999. Does aluminium affect root growth of maize through
interaction with the cell wall–plasma membrane–cytoskeleton
continuum? Plant and Soil 215, 163–174.
Hossain AKMZ, Koyama H, Hara T. 2006. Growth and cell wall
properties of two wheat cultivars differing in their sensitivity to
aluminum stress. Journal of Plant Physiology 163, 39–47.
Huang Y, Man H-Y, Sekine-Aizawa Y, Han Y, Juluri K, Luo H,
Cheah J, Lowenstein C, Huganir RL, Snyder SH. 2005. S-
nitrosylation of N-ethylmaleimide sensitive factor mediates surface
expression of AMPA receptors. Neuron 46, 533–540.
Ishikawa H, Evans ML. 1993. The role of the distal elongation
zone in the response of maize roots to auxin and gravity. Plant
Physiology 102, 1203–1210.
Jones DL, Blancaflor EB, Kochian LV, Gilroy S. 2006. Spatial
coordination of aluminium uptake, production of reactive oxygen
species, callose production, and wall rigidification in maize roots.
Plant, Cell and Environment 29, 1309–1318.
Kakegawa W, Yuzaki M. 2005. A mechanism underlying AMPA
receptor trafficking during cerebellar long-term potentiation.
Proceedings of the National Academy of Sciences USA 102,
17846–17851.
Klimashevski EL, Dedov VM. 1975. Localization of the mechanism
of growth inhibiting action of Al 3+ in elongating cell walls.
Fiziologija Rastenij 22, 1040–1046.
Kollmeier M, Felle HH, Horst WJ. 2000. Genotypical differences
in aluminum resistance of maize are expressed in the distal part of
the transition zone. Is reduced basipetal auxin flow involved in
inhibition of root elongation by aluminum? Plant Physiology 122,
945–956.
Lazof DB, Goldsmith JG, Rufty TW, Linton RW. 1994. Rapid
uptake of aluminum into cells of intact soybean root tips. Plant
Physiology 106, 1107–1114.
Lazof DB, Goldsmith JG, Rufty TW, Linton RW. 1996. The early
entry of Al into cells of soybean roots. Plant Physiology 112,
1289–1300.
Lombardo MC, Graziano M, Polacco JC, Lamattina L. 2006.
Nitric oxide functions as a positive regulator of root hair
development. Plant Signaling and Behavior 1, 28–33.
Mancuso S, Marras AM, Volker M, Baluška F. 2005. Noninvasive
and continuous recordings of auxin fluxes in intact
root apex with a carbon-nanotube-modified and self-referencing
microelectrode. Analytical Biochemistry 341, 344–351.
Marienfeld S, Schmohl N, Klein M, Schröder WH, Kuhn AJ,
Horst W. 2000. Localization of aluminium in root tips of Zea mays
and Vicia faba. Journal of Plant Physiology 156, 666–671.
Miyasaka SC, Kochian LV, Shaff JE, Foy CD. 1989. Mechanism
of aluminium tolerance in wheat. An investigation of genotype
differences in rhizosphere pH, K, and H transport, and root-cell
membrane potentials. Plant Physiology 91, 1188–1196.
Meffert MK, Calakos NC, Scheller RH, Schulman H. 1996.
Nitric oxide modulates synaptic vesicle docking/fusion reactions.
Neuron 16, 1229–1236.
Murashige T, Skoog F. 1962. A revised medium for rapid growth
and biosynthesis with tobacco tissue culture. Physiologia Plantarum
15, 473–497.
Nagy NE, Dalen LS, Jones DL, Swensen B, Fossdal CG,
Eldhuset TD. 2004. Cytological and enzymatic responses to
aluminium stress in root tips of Norway spruce seedlings. New
Phytologist 163, 595–607.
Olivetti PG, Cumming JR, Etherton B. 1995. Membrane potential
depolarization of root cap cells precedes aluminum tolerance in
snapbean. Plant Physiology 109, 123–129.
Ovečka M, Lang I, Baluška F, Ismail A, Illéš P, Lichtscheidl IK.
2005. Endocytosis and vesicle trafficking during tip growth of root
hairs. Protoplasma 226, 39–54.
Paciorek T, Zažímalová E, Ruthhardt N, et al. 2005. Auxin
inhibits endocytosis and promotes its own efflux from cells.
Nature 435, 1251–1256.
Pavlovkin J, Mistrík I.1999. Phytotoxic effect of aluminium on
maize root membranes. Biologia 54, 473–479.
Pavlovkin J, Mistrík I, Zajchenko AM, Dugovič L. 1993.
Some aspects of phytotoxic action of trichothecene mycotoxin
roridin H on corn roots. Biologia 48, 435–439.
Polle E, Konzak CF, Littrik JA. 1978. Visual detection of
aluminium tolerance levels in wheat by haematoxylin staining of
seedling roots. Crop Science 18, 823–827.
Ryan PR, DiTomaso JM, Kochian LV. 1993. Aluminium toxicity in
roots: an investigation of spatial sensitivity and the role of the root
cap. Journal of Experimental Botany 44, 437–446.
Šamaj J, Baluška F, Voigt B, Schlicht M, Volkmann D,
Menzel D. 2004. Endocytosis, actin cytoskeleton, and signaling.
Plant Physiology 135, 1150–1161.
Šamaj J, Read ND, Volkmann D, Menzel D, Baluška F. 2005. The
endocytic network in plants. Trends in Cell Biology 15, 425–433.

Aluminium sensitivity of root cells related to aluminium internalization 4213
Santelia D, Vincenzetti V, Azzarello E, Bovet L, Fukao Y,
Düchtig P, Mancuso S, Martinoia E, Geisler M. 2005. MDRlike
ABC transporter AtPGP4 is involved in auxin-mediated lateral
root and root hair development. FEBS Letters 579, 5399–5406.
Schlicht M, Strnad M, Scanlon MJ, Mancuso S, Hochholdinger
F, Palme K, Volkmann D, Menzel D, Baluška F. 2006. Auxin
immunolocalization implicates vesicular neurotransmitter-like
mode of polar auxin transport in root apices. Plant Signaling and
Behavior 1, 122–133.
Schmohl N, Horst WJ. 2000. Cell wall pectin content modulates
aluminium sensitivity of Zea mays (L.) cells grown in suspension
culture. Plant, Cell and Environment 23, 735–742.
Silva IR, Smyth TJ, Moxley DF, Carter TE, Thomas EC, Allen
NS, Rufty TW. 2000. Aluminum accumulation at nuclei of cells
in the root tip. Fluorescence detection using lumogallion and confocal
laser scanning microscopy. Plant Physiology 123, 543–552.
Sivaguru M, Baluška F, Volkmann D, Felle HH, Horst WJ. 1999.
Impact of aluminum on the cytoskeleton of the maize root apex.
Short-term effects on the distal part of the transition zone. Plant
Physiology 119, 1073–1082.
Sivaguru M, Ezaki B, He ZH, Tong H, Osawa H, Baluška F,
Volkmann D, Matsumoto H. 2003a. Aluminum-induced gene
expression and protein localization of a cell wall-associated receptor
kinase in Arabidopsis. Plant Physiology 132, 2256–2266.
Sivaguru M, Horst WJ. 1998. The distal part of the transition zone
is the most aluminium-sensitive apical root zone of Zea mays L.
Plant Physiology 116, 155–163.
Sivaguru M, Pike S, Gassmann W, Baskin TI. 2003b. Aluminum
rapidly depolymerizes cortical microtubules and depolarizes the
plasma membrane: evidence that these responses are mediated by
a glutamate receptor. Plant Cell Physiology 44, 667–675.
Taylor GJ, McDonald-Stephens JL, Hunter DB, Bertsch DB,
Elmore D, Rengel Z, Reid RJ. 2000. Direct measurement of
aluminum uptake and distribution in single cells of Chara
corallina. Plant Physiology 123, 987–996.
Tice KR, Parker DR, DeMason DA. 1992. Operationally defined
apoplastic and symplastic aluminum fractions in root tips of
aluminum-intoxicated wheat. Plant Physiology 100, 109–318.
Vázquez MD. 2002. Aluminum exclusion mechanism in root tip of
maize (Zea mays L.): lysigeny of aluminum hyperaccumulator
cells. Plant Biology 4, 234–249.
Vázquez MD, Poschenrieder C, Corrales I, Barceló J. 1999.
Change in apoplastic aluminum during the initial growth response
to aluminum by roots of tolerant maize variety. Plant Physiology
119, 435–444.
Verbelen J-P, De Cnodder T, Le J, Vissenberg K, Baluška F.
2006. The root apex of Arabidopsis thaliana consists of four
distinct zones of cellular activities. Meristematic zone, transition
zone, fast elongation zone, and growth terminating zone. Plant
Signaling and Behavior (in press).
Vida TA, Emr SD. 1995. A new vital stain for visualizing vacuolar
membrane dynamics and endocytosis in yeast. Journal of Cell
Biology 128, 779–792.
Vitorello VA, Haug A. 1996. Short-term aluminium uptake by
tobacco cells: growth dependence and evidence for internalization
in a discrete peripheral region. Physiologia Plantarum 97,
536–544.
Vitorello VA, Haug A. 1997. An aluminium-morin fluorescence
assay for the visualization and determination of aluminium in
cultured cells of Nicotiana tabacum L. BY-2. Plant Science 122,
35–42.
Voigt B, Timmers ACJ, Šamaj J, et al. 2005. Actin-based motility
of endosomes is linked to the polar tip-growth of root hairs.
European Journal of Cell Biology 84, 609–621.
Wang Y, Stass A, Horst WJ. 2004. Apoplastic binding of aluminum
is involved in silicon-induced amelioration of aluminum toxicity in
maize. Plant Physiology 136, 3762–3770.
Wang G, Moniri NH, Ozawa K, Stamler JS, Daaka Y. 2006.
Nitric oxide regulates endocytosis by S-nitrosylation of dynamin.
Proceedings of the National Academy of Sciences, USA 103,
1295–1300.
Yu Q, Hlavačka A, Matoh T, Volkmann D, Menzel D, Goldbach
HE, Baluška F. 2002. Short-term boron deprivation inhibits
endocytosis of cell wall pectins in meristematic cells of maize and
wheat root apices. Plant Physiology 130, 415–421.