Journal of Experimental Botany, Vol. 57, No. 9, pp. 2061–2074, 2006
doi:10.1093/jxb/erj159 Advance Access publication 23 May, 2006
RESEARCH PAPER
Localization of arabinogalactan proteins in egg cells,
zygotes, and two-celled proembryos and effects of
b-D-glucosyl Yariv reagent on egg cell fertilization and
zygote division in Nicotiana tabacum L.
Yuan Qin and Jie Zhao*
Key Laboratory of the Ministry of Education for Plant Developmental Biology, College of Life Sciences,
Wuhan University, Wuhan 430072, China
Received 13 September 2005; Accepted 8 February 2006
Abstract
Arabinogalactan proteins (AGPs) have been implicated
in a variety of plant development processes including
sexual plant reproduction. As a crucial developmental
event, plant sexual reproduction generally occurs inside an ovule embedded in an ovary. The inaccessibility of the egg cells, zygotes, and embryos has hindered
our understanding of the importance of AGPs in the
early events involving fertilization, zygotic division, and
early embryogenesis. In this study, the well-established
in vitro zygote and ovary culture systems, together
with immunofluorescence and immunogold labelling
techniques, were employed to investigate the role of
AGPs in the early events of sexual reproduction in
Nicotiana tabacum. Dramatic changes in AGP content
during ovule development were evidenced by western
blotting. Subcellular localization revealed that AGPs
are localized in the plasma membrane, cell wall, and
cytoplasm of pre- and post-fertilized egg cells, and
cytoplasm and vacuoles of two-celled proembryos.
Abundant AGPs were detected in unfertilized egg cells;
however, the level of AGPs substantially decreased in
fertilized egg cells. Polar distribution of AGPs in elongated zygotes was observed. The early two-celled
proembryos just from zygote division displayed accumulation of AGPs at a low level, while in the
elongated two-celled proembryos at the late stage,
the AGP content clearly increased. Provision of
bGlcY, a synthetic phenylglycoside that specifically
binds AGPs, to the in vitro cultures of isolated zygote
and fertilized ovaries increased abnormal symmetrical
division of zygotes. In the culture of pollinated but
unfertilized ovaries, addition of bGlcY resulted in
arrest of fertilization of the egg cells, but had no
effect on fertilization of the central cells. The possible
roles of AGPs in fertilization, zygotic division, and
proembryo development are discussed.
Key words: Arabinogalactan protein, egg cell, monoclonal
antibody, Nicotiana tabacum L., Yariv reagent, zygote
Introduction
Arabinogalactan proteins (AGPs) are a diverse family of
hydroxyproline-rich glycoproteins (HRGPs). AGPs are
widely distributed in higher plants and play multiple roles
in various processes associated with plant growth and development, including cell expansion, cell proliferation, and
embryogenesis (Zhu et al., 1993; Knox, 1996; Nothnagel,
1997). AGPs typically contain >90% (w/w) carbohydrates
covalently attached to a core protein backbone that is
usually enriched in hydroxyproline, alanine, serine, and
threonine. The polysaccharide is composed of b-1,3- and
b-1,6-linked galactans containing uronides and neutral
sugars including arabinose (Fincher et al., 1983). Several
monoclonal antibodies against the carbohydrate epitopes of
AGPs have been used extensively to examine AGP expression and its potential functions (Knox, 1997). Amongst
them, Mb JIM4, JIM13, and LM2 recognize different
carbohydrate epitopes of AGPs. Mb JIM4 and JIM13
react with the b-D-GlcpA-(1!3)-a-D-GalpA-(1!2)-L-Rha
* To whom correspondence should be addressed. E-mail: [email protected]
Abbreviations: AGP, arabinogalactan-protein; bGlcY, b-glucosyl Yariv regent; DAP, days after pollination; PBS, phosphate-buffered saline.
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: [email protected]
2062 Qin and Zhao
epitope, but LM2 reacts with the b-D-GlcpA epitope
(Showalter, 2001). In addition, the use of Yariv reagent
(bGlcY), a b-D-glucosyl derivative of phloroglucinol that
can specifically bind to AGPs to perturb AGP functions in
living cells and seedlings, has suggested roles for AGPs
in cell growth and development (Majewska-Sawka and
Nothnagel, 2000). The studies using monoclonal antibodies
and Yariv reagent have generated valuable information on
the tissue- and cell-specific expression patterns of AGPs.
More recently, a mutant, designated as sos5, showing hypersensitivity to NaCl, abnormal cell expansion, and reduced
seed number was identified in a forward genetic screen
(Shi et al., 2003). SOS5 encodes an AGP-like protein.
Besides its important role in cell expansion and cell–cell
adhesion, SOS5 might also play roles in plant reproduction (Shi et al., 2003).
Plant sexual reproduction is an extraordinarily complicated process including several key events: pollination,
fertilization, and embryogenesis. The presence of AGPs
in the reproductive tissues and its alternation during
sexual reproduction suggested involvement of AGPs in
this complex process (Wu et al., 2001). The use of antibody
MAC207 has shown an alteration of cell surface epitope
expression during sexual development in Pisum sativum
(Pennell and Roberts, 1990). Changes in the temporal and
spatial expression of AGP during sexual reproduction
have also been demonstrated in oilseed rape by detection
of the plasma membrane AGP using Mb JIM8 (Pennell
et al., 1991).
Although the molecular mechanism of fertilization in
higher plants is still unclear, it is believed that successful
fertilization depends on recognition and interaction between male and female gametophytes. Evidence has shown
that successful recognition and fusion between egg and
sperm require the specificity of gamete cell surfaces (Knox
et al., 1988). Faure et al. (1994) reported that the frequency
of in vitro fusion between sperm and egg cells can reach
80%, which is much higher than that of the fusion between
sperm cell and mesophyll (2%), and between two sperm
cells (18%). This further supports the notion that fertilization
requires recognition between sperm and egg cell through
their cell surface molecules. The presence of JIM8 and
JIM13 epitopes on the plasma membrane of sperm and
vegetative cells suggests a role for AGPs in sperm–
egg cell recognition (Russell, 1985; Southworth and
Kwiatkowski, 1996). The differential expression of the
JIM8 epitope in the paired sperm cells, egg cell, and
central cell of oilseed rape imply that different AGPs
might function differently in sperm–egg recognition
and sperm–central cell recognition during the double
fertilization of angiosperms (Pennell et al., 1991).
AGPs have also been implicated in somatic embryogenesis. The correlation of the presence of the JIM4 epitopes
with certain stages of somatic embryogenesis supports
a role for AGPs in embryo growth and differentiation
(Stacey et al., 1990). Moreover, the AGPs secreted from
culture cells or extracted from seeds affect the induction
and development of somatic embryos (Kreuger and van
Holst, 1993; Egertsdotter and von Arnold, 1995). Since
bGlcY can specifically bind AGP and perturb its biological activity, bGlcY has also been used to explore the
role of AGPs in embryogenesis. Embryogenic carrot cultures grown in the presence of bGlcY could not produce
normal embryos (Thompson and Knox, 1998). The effect
of bGlcY on inhibition of Cichorium somatic embryogenesis was concentration-dependent and reversible
(Chapman et al., 2000).
Because egg cells and zygotes reside deep inside various
sporophytic tissues of ovules and ovaries, it has been
technically difficult to explore the functions of AGPs in
fertilization of egg cells and division of zygotes. Indeed,
the roles of AGPs and bGlcY in egg cell fertilization and
zygote division in angiosperms is obscure. In these studies,
the well-established tobacco egg cell and zygote isolation
and in vitro microculture system was used to tackle the
difficulty of accessing these sexual cells in planta. In
combination with immunofluorescence and immunogold
labelling, the temporal and spacial expression and subcellular localization of AGPs in sexual cells and proembryos were detected. The biological functions of AGPs
were also investigated by examining the effects of bGlcY
on egg cell fertilization and zygote division.
Materials and methods
Plant materials
Nicotiana tabacum L. (cv. Xanthi and SR1) plants were grown under
standard greenhouse conditions. Flowers were artificially pollinated
during anthesis.
Total protein extraction from ovules
Ovular proteins were extracted as described by Wittink et al. (2000)
with minor modifications. Briefly, 1 g (fresh weight) of ovules at
differential developmental stages (1, 3, 5, 6, 7, 8, and 9 d after
pollination, DAP) isolated from the placenta by dissection were
ground to a fine powder in liquid nitrogen. A 2 ml aliquot of
extraction buffer (0.1 M K3PO4, pH 7.0) was added to the ground
tissue and mixed. After incubation at 4 8C for 3 h, the mixture was
centrifuged at 12 000 rpm for 20 min. The supernatant was
precipitated with 5 vols of acetone overnight at 20 8C. The pellet
was resuspended in 0.25 ml of 50 mM Tris-HCl, pH 8.0. After
centrifugation, the insoluble material was subjected to a second
round of extraction by using 0.25 ml of 50 mM Tris-HCl, pH 8.0.
The supernatants from both extractions were pooled and stored at
70 8C until analysis.
Analysis by SDS–PAGE and immunoblot assays
The soluble proteins from ovules at different stages were separated
by SDS–PAGE using a 12% acrylamide separating gel and a 5%
acrylamide stacking gel in a Mini-Protean II electrophoresis cell
(Bio-Rad). After electrophoresis, proteins were transferred onto
a nitrocellulose membrane by electroblotting at 88 V for 3 h using
electrotransfer buffer (20 mM TRIS-base, 150 mM glycine, 20%
methanol). The nitrocellulose membrane blots were blocked with
Egg cell fertilization and zygote division in N. tabacum L.
5% non-fat dried milk in TTBS buffer (20 mM TRIS-base, 500 mM
NaCl, 0.05% Tween-20, pH 7.5) for 1 h and then incubated at room
temperature with the primary antibody (1:100) against the AGP
epitope for 2 h. After three washes with TTBS, the blots were
incubated with alkaline phosphatase-conjugated goat anti-rat antibody (1:500; Sino-American Biotechnology Co.) for 1 h at room
temperature followed by three washes with TTBS buffer. The
alkaline phosphatase signal was developed using the nitro
blue tetrazolium/5-bromo-4-chloro-3-indolyl phosphate kit (SinoAmerican Biotechnology Co.). The pre-stained protein mixture
(Pierce) containing myosin (223 kDa), phosphorylase B (110 kDa),
bovine serum albumin (BSA; 81.6 kDa), ovalbumin (46.8 kDa),
carbonic anhydrase (31.8 kDa), trypsin inhibitor (24.8 kDa), and
lysozyme (16.5 kDa) was used as the molecular weight marker.
Isolation of embryo sacs
Tobacco ovules isolated by dissection (1 DAP) were fixed in FPA
solution (formaldehyde solution:propionic acid:70% ethanol, 5:5:90)
overnight at 4 8C. The fixed ovules were rinsed with 70% ethanol 2–3
times and then permeated by a decreasing series of ethanol solution
through to water. The ovules were incubated in enzyme solution
(2% cellulose R-10 and 1.5% macerozyme R-10, pH 5.7) for 3 h at
30 8C with shaking. After three washes with water to remove enzymes, the embryo sacs were released from ovules by pipetting
and collected using a micropipette.
Isolation of egg cells, synergids and central cells
The isolation process was performed as described by Tian and
Russell (1997). Tobacco ovules were dissected from ovaries (1 DAP)
and placed into enzyme solution containing 13% mannitol, 3 mM
2-[N-morpholino]ethanesulphonic acid (MES), 3 mM polyvinypyrrolidone K30 (PVP K30), 1% cellulose R-10, and 0.8% macerozyme
R-10, pH 5.7, then incubated with vibration for 3 h on an oscillator
(ZW-A, FU-HUA). After washing three times, egg cells, synergids,
and central cells were released from ovules by pipetting and collected
with a micropipette.
Isolation of zygotes and proembryos
Isolation of zygotes was performed according to the methods reported
previously (Li and Yang, 2000). Fertilized embryo sacs were isolated
by enzymatic maceration of ovules for 10–30 min at 25 8C in the
solution containing 13% mannitol, 3 mM MES, 3 mM PVP K30, 1%
cellulose R-10, and 0.8% macerozyme R-10, pH 5.7, and then
gently grinding the macerated ovules with a small glass pestle. The
zygotes or proembryos were isolated from embryo sacs by microdissection after a short-term enzymatic treatment of embryo sacs with
solution containing 13% mannitol, 3 mM MES, 3 mM PVP K30,
0.25% cellulose R-10, and 0.2% macerozyme R-10, pH 5.7.
Immunolocalization of AGPs by light microscopy
The samples were fixed for 1 h at room temperature in 1.5%
paraformaldehyde in 50 mM PIPES buffer pH 6.7, 2 mM MgSO4.
7H2O, 2 mM EGTA, 13% mannitol. After three rinses with 50 mM
PIPES buffer (pH 6.7) containing 2 mM MgSO4. 7H2O and 2 mM
EGTA and one rinse with 100 mM phosphate-buffered saline (PBS),
pH 7.4, the fixed samples were incubated in the primary Mb JIM13
or LM2 diluted 1:10 in 100 mM PBS (pH 7.4) for 2 h at room
temperature. After three rinses with 100 mM PBS (pH 7.4), the
samples were then incubated for 1 h in the dark with the secondary
antibody, anti-rat-IgG–fluorescein isothiocyanate (FITC) conjugate
(Sigma) diluted 1:100 with 100 mM PBS (pH 7.4). The samples were
rinsed three times with 100 mM PBS (pH 7.4) before microscopic
examination (Zhao et al., 2004). In the control, the samples were
incubated in 100 mM PBS (pH 7.4) instead of the primary antibody.
2063
Immunolocalization of AGPs by transmission electron
microscopy (TEM)
The isolated ovules were fixed in a mixture of 3% paraformaldehyde
and 1% glutaraldehyde in 10 mM PBS, pH 7.2 under vacuum for 2 h
at room temperature and then kept in fresh fixatives at 4 8C overnight.
After fixation, the samples were rinsed three times with 10 mM PBS
(pH 7.2), and then dehydrated gradually in a series of increasing
concentrations of ethanol for 30 min for each step at 10, 30, and 50%
(at 4 8C), and 70, 90, and 95% (at 20 8C). In the last two steps
of dehydration in 100% absolute ethanol, the samples were treated
for 60 min. The samples were infiltrated with an increasing ratio of
Lowicryl K4M:ethanol at 1:2, 1:1, and 2:1 for 12 h at each step
followed by Lowicryl K4M without ethanol for 12 h at 20 8C. The
infiltration was completed with fresh Lowicryl K4M for 1 day at 20
8C. The samples were transferred into capsules containing fresh
Lowicryl K4M, and cured under two 15 W ultraviolet lamps (360 nm)
for at least 24 h at 20 8C, and then curing was continued under
UV light for 2 d at room temperature.
Ultrathin sections (60 nm) were prepared using a Sorvall MT-6000
ultramicrotome and collected onto Formvar-coated nickel grids. The
sections were incubated in PBST buffer (60 mM PBS, 0.1% Tween20, 0.02% NaN3, pH 7.2) containing 0.2 M glycine and 1% BSA for
20 min to block non-specific binding and then incubated with AGP
antibody JIM13 at 37 8C for 3 h. The sections were rinsed three times
with PBST, and then incubated with goat anti-rat IgG conjugated to
10 nm gold particles (Sigma) at 1:100 dilution for 1 h at 37 8C. After
washing three times with PBST and three times with ddH2O, the
sections on the grids were air-dried and post-stained with saturated
uranyl acetate for 15 min and then with lead citrate for 15 min.
Control sections were treated similarly except that primary antibody
was omitted. The sections were examined and photographed under
a JEM 100/II TEM.
Zygote culture
Ovaries at 3–3.5 DAP were sterilized with 70% ethanol for 0.5–1 min
and then with 2% NaOCl for 4 min followed by three washes with
sterile distilled water. Isolation of embryo sacs and zygotes is
essentially the same as described above but under sterile conditions.
Isolated zygotes were washed with 13% mannitol once and with
medium twice. The medium is composed of the MS main salt
plus KM8p minimal salt, organic elements, and vitamins (MsK)
supplemented with 0.05 mol l1 glucose, 0.25 mol l1 sucrose,
0.15 mol l1 mannitol, 0.1 mol l1 sorbitol, 1 mg l1 2,4-D, 0.5 mg
l1 6-BA, and 10% coconut, pH 5.8. After washing, the isolated
zygotes were transferred into a millicell with a semi-permeable
membrane (MILLICELL-CM 0.4 lm PICMO 1250, diameter 12
mm) placed in the middle of a 35 mm Petri dish with 1.5 ml of
medium containing 100–120 ovules (4 DAP) as feeder cells residing
outside the minicell. For bGlcY treatment, bGlcY (Biosupplies
Pty, Australia) was added to the medium at final concentrations of
10, 30, 50, 80, and 100 lM. The zygotes were cultured in the
dark at 25 8C. After 2–3 d of culture, the frequency and pattern
of zygote division were scored. Calcofluor white ST (CW) was employed to examine the cell wall formation of zygotes. The samples
were examined under an inverted microscope (Olympus IMT-2)
and photographed with a CoolSNAP CCD (Photometrics, USA).
Ovary culture
Ovaries at 1.5 and 3 DAP were sterilized with 70% ethanol for
0.5–1 min and then with 2% NaOCl for 4 min. After rinsing 3–4 times
with sterile distilled water, the ovaries were cultured in MsK
medium supplemented with 0.03 mol l1 glucose, 0.15 mol l1
sucrose, 0.1 mol l1 mannitol, 0.6 mol l1 sorbitol, 1 mg l1 2,4-D,
0.5 mg l1 6-BA, and 10% coconut, pH 5.8 by immersing the flower
2064 Qin and Zhao
stalks of ovaries in the medium. bGlcY at 10, 50, and 100 lM and
100 lM aGalY (a Yariv reagent incapable of binding AGPs) were
added to the medium. The ovaries were cultured under 14 h daylight
and 10 h darkness at 25 8C. After 4 d, the lengths and widths of
ovaries and ovules were measured. For microscopic observation,
some ovules were fixed and dehydrated as described previously,
and then infiltrated with 1,2-epoxypropane and embedded with
Epon 812 resin. Some ovules were placed into enzyme solution for
isolation of embryos. The samples were examined under a Leica
DM IRB inverted microscope and photographed with a CooledCCD (MicroMax Princeton Instruments, Inc.) or a CoolSNAP CCD
(Photometrics, USA). Experiments were repeated at least five
times run in 4–5 replicates each time. The total number of ovules
counted in each group was between 100 and 120 each time, and
the P-values were calculated according to the untreated series as
controls.
Quantitative analysis of cell size
The quantitative analysis of cell size was combined with statistical
analysis. Five or six ovules from variously treated ovary cultures
were sampled for sectioning, and three or four sections from each
ovule were analysed. The sections were examined under a Leica DM
IRB inverted microscope and photographed with a Cooled-CCD
(MicroMax Princeton Instruments, Inc.). The areas of the nucellar
cells were counted with MetaMorph Software (MicroMax Princeton
Instruments, Inc.).
Results
Expression of AGPs during tobacco ovule
development
JIM4, JIM13, and LM2 are rat monoclonal antibodies
that recognize structurally complex carbohydrate epitopes
found on different AGP core proteins (Showalter, 2001).
To examine the expression of different kinds of AGPs
in tobacco ovules, the proteins from 5–8 DAP tobacco
ovules were extracted, separated by SDS–PAGE, electroblotted onto nitrocellulose membranes, and immunoblotted with Mb JIM4, JIM13, and LM2. The AGPs which
reacted with Mb LM2 appear to be broad smears with
apparent molecular weights of 90–223 kDa (Fig. 1A).
In contrast, the JIM13-labelled AGPs ranged from 40 to
223 kDa in size and were more abundant than those reacting with LM2 (Fig. 1A). These smeared bands were a
characteristic of AGPs. Although JIM13 and JIM4 recognize the same epitope in AGPs, the AGPs recognized by
JIM4 were hardly detectable in tobacco ovules in this
study (Fig. 1A). Therefore, only JIM13 and LM2 antibodies
were used for the following studies.
To detect differences in the expression of AGPs during
tobacco ovule development, the extracts from ovules at
different developmental stages were analysed by western
blotting. Figure 1B shows that both the abundance and
molecular weight of LM2-labelled AGPs were changed
dramatically during tobacco ovule development. Abundant
AGPs ranging from 70 to 223 kDa were detected in
unfertilized ovules (1 DAP). However, the content of the
Fig. 1. Western blot analysis of AGPs recognized by Mb JIM4, JIM13,
and LM2. An equal amount of total proteins was loaded and separated by
12.5% SDS–PAGE. (A) Expression of JIM13 (lane 1), LM2 (lane 2), and
JIM4 (lane 3) epitopes in tobacco ovules 5–8 d after pollination (DAP).
(B) Expression of LM2 epitopes in the ovules at 1–9 DAP. (C)
Expression of JIM13 epitopes in the ovules at 1–9 DAP. Molecular mass
(kDa) is indicated on the left.
LM2 epitope declined quickly in the ovules in which the
egg cells are fertilized (3 DAP). A low level of LM2
epitope was present in the ovules at the early (5 DAP) and
middle (6 DAP) stages of the globular proembryo, while
the expression of AGPs in the ovules at the late stage of
the globular proembryo (7 DAP) and the heart-shaped
embryo (8 DAP) increased gradually until the stage of the
torpedo-shaped embryo (9 DAP). Compared with the
expression of AGPs recognized by LM2, the expression
Egg cell fertilization and zygote division in N. tabacum L.
2065
Fig. 2. Localization of AGPs by immunofluorescence labelling. A bright-field micrograph and its fluorescent image are shown in each panel. Cells in
(A–F) and (H) were labelled by JIM13, and that in (G) was labelled by LM2. (A) An isolated embryo sac before fertilization showing strong fluorescence
at the micropylar end (arrowhead). (B) A mature egg cell showing uniform and strong fluorescence on the cell surface. (C) A synergid cell showing
strong fluorescent labelling on the cell surface. (D) A central cell displaying weak fluorescence labelling. (E) A zygote at the early stage showing weak
fluorescence labelling. (F) An elongated zygote displaying polar distribution of fluorescence. (G) An elongated zygote labelled by LM2. Fluorescence
was concentrated in the region near the zygote nucleus (chalazal region). (H) A two-celled proembryo at the late stage having increased fluorescence
compared with zygotes. Bar=10 lm.
of the JIM13 epitope in the ovules at different stages was
less changed and showed a relatively broad molecular
weight range (Fig. 1C). AGPs recognized by JIM13 with
molecular weights from 70 to 223 kDa displayed higher
expression in the ovules at all tested developmental stages.
The changes of AGPs during ovule development, in
particular before and after fertilization, suggest roles for
AGP in plant sexual reproduction.
Immunolocalization of AGPs in isolated embryo sacs,
egg cells, synergid cells, central cells, zygotes, and
two-celled proembryos
In order to examine the distribution of AGPs in embryo
sac and viable female cells, immunolocalization with Mb
JIM13 was performed at the light microscope level. As
shown in Fig. 2A, the isolated embryo sac before
fertilization showed strong fluorescence labelling in the
2066 Qin and Zhao
egg apparatus at the mycropylar end (arrowhead) and in
antipodal cells at the chalazal end of isolated embryo sacs,
while weak fluorescence was observed in the embryo sac
wall and central cell. This result indicates that AGPs are
differentially expressed in the female cells within embryo
sacs. The distribution of AGPs was investigated further
by immunofluorescent labelling of isolated mature unfertilized egg cells, synergid cells, and central cells with
JIM13 and LM2 (data not shown). A strong and uniform
fluorescent ring along the plasma membrane of egg cells
(Fig. 2B) and synergid cells (Fig. 2C) was observed,
while weak fluorescence was present on the plasma membrane of isolated central cells (Fig. 2D). Interestingly, the
zygotes (fertilized egg cells) displayed poor fluorescence
with both JIM13 (Fig. 2E) and LM2 (data not shown). A
polar distribution pattern of JIM13-reactive AGPs in the
elongated zygotes (3–3.5 DAP) was observed (Fig. 2F).
Stronger fluorescence in the area near the zygote nucleus
(chalazal region) was also detected by LM2 labelling
(Fig. 2G). After the first zygotic division, the two-celled
proembryos at the early stage displayed weak and uniform
fluorescence (data not shown). However, the intensity of
fluorescence labelling by JIM13 in the late two-celled
proembryos was increased (Fig. 2H). The control cells not
labelled with JIM13 or LM2 showed no fluorescence
(data not shown).
To detect the subcellular localization of AGPs, immunogold labelling and TEM techniques were used. These
techniques enable precise localization of AGPs in subcellular compartments within cells. Using Mb JIM13,
abundant gold particles were detected on the plasma
membrane and in the cytoplasm of pre-fertilized egg cells
(Fig. 3A), while the control sections without incubation
with Mb JIM13 showed no gold particle (Fig. 3B). In
contrast, only a few scattered gold particles were detected in
the cytoplasm and vacuoles of fertilized egg cells at the
stage when the egg nucleus has not fused with the sperm
nucleus (Fig. 3C). The JIM13 epitopes were sparsely
labelled in the zygotes (data not shown) and newly formed
two-celled proembryo (Fig. 4A, B). However, increased
gold particles were observed in the elongated two-celled
proembryo at the late stage (Fig. 4C, D). Intriguingly, the
gold particles often appeared in the cytoplasm and vacuoles
rather than the plasma membrane and cell wall in twocelled proembryos.
Effects of bGlcY on zygotic division during isolated
zygote culture
For efficient isolation of viable zygotes, enzymatic maceration was combined with a microdissection technique. To
obtain zygotes with an intact cell wall, isolated embryo sacs
were subjected to enzymatic maceration for ;2–10 min
depending on the integrity of the isolated embryo sac, and
then zygotes were isolated from these embryo sacs by
Fig. 3. TEM immunolocalization of the AGP epitope recognized by
JIM13 in an egg cell. (A) Dense gold particles (arrowheads) were present
in the cytoplasm and plasma membrane of a pre-fertilized egg cell at
1 DAP. (B) No gold particles were detected in an egg cell without
primary antibody incubation as a control test. (C) At 2 DAP, a sperm cell
penetrated the egg cell and only scattered gold particles (arrowheads)
were detected in the egg cell and sperm cell. CC, central cell; EC, egg cell,
EN, egg cell nucleus; SN, sperm nucleus. Bar=1 lm.
microdissection with glass microneedles (Fig. 5A, B). The
zygotes underwent first division after 5–15 h in vitro
culture. Most elongated zygotes (Fig. 5B) divided asymmetrically to form two-celled proembryos consisting of
a small apical cell and a large basal cell (Fig. 5C), which
resembles the natural division pattern of zygotes during
sexual reproduction in tobacco. However, symmetrical
divisions (Fig. 5D) were also observed in some spherical
zygotes (Fig. 5A), which are either non-polarized zygotes
at an early stage or zygote protoplasts formed by excessive enzymatic treatment. After 2–3 d of culture, multicellular proembryos (Fig. 5E) were formed in the medium
with ovules as feeder tissues and the presence of the cell
wall was evidenced by CW fluorescence (Fig. 5F).
The effects of bGlcY on zygotic division and proembryo
formation were studied by addition of different concentrations of bGlcY to the culture medium. As depicted in
Table 1, the frequency and pattern of the first zygotic division were not significantly influenced by addition of 10 lM
bGlcY. However, when treated with 50 lM bGlcY, nearly
one-third of the zygotes divided symmetrically (Fig. 5G)
and 22.9% displayed budded division (Fig. 5H), while
Egg cell fertilization and zygote division in N. tabacum L.
2067
Fig. 4. TEM immunolocalization of an AGP epitope recognized by JIM13 in two-celled proembryos. (A) An ultrathin section of an early two-celled
proembryo. (B) An enlarged image of the area indicated by a square in (A). A few gold particles (arrowheads) were detected in the cytoplasm of the early
two-celled proembryo. (C) Ultrathin section of a late two-celled proembryo. (D) An enlarged image of the area indicated by a square in (C). An increased
number of gold particles (arrowheads) appeared in the cytoplasm of the late two-celled proembryo. AC, apical cell; BC, basal cell; C, cytoplasm; CW,
cell wall; M, mitochondrion; PM, plasma membrane; V, vacuole. Bar=1 lm.
normal asymmetric division (Fig. 5I) was reduced to
31.2%. When the concentration of bGlcY increased to
80 lM in the medium, the frequency of symmetric zygote
division increased to 37.8%. Under treatment with this
concentration, two-celled proembryos (Fig. 5J) resulting
from symmetric division of zygotes could continue to
divide and form a multicellular proembryo-like structure
without a typical embryo proper and suspensor (Fig. 5K).
Interestingly, asymmetrically divided zygotes (Fig. 5L)
could still develop into proembryos with the typical
structure of an embryo body and suspensor in the presence
of 80 lM bGlcY in the culture medium (Fig. 5M). The
frequency and pattern of the first zygotic division under
treatment with 100 lM bGlcY were similar to that with
80 lM bGlcY treatment. Symmetric (Fig. 5N) and asymmetric (Fig. 5O) divisions could still be observed at this
high concentration, and asymmetrically divided zygotes
could also develop into four-cell proembryos (Fig. 5P).
Effects of bGlcY on egg cell fertilization and zygote
division during ovary culture
By examining pollen tube growth, fertilization, and early
embryogenesis within ovaries, it was established that 1.5
DAP ovaries are at the stage when pollen tubes reach the
bottom of styles and egg cells are not fertilized, and 3 DAP
ovaries are at the stage when egg cells are fertilized to form
zygotes. The ovaries at these two stages with a total number of ;25 at each stage were chosen for in vitro culture.
After 4 d of in vitro culture of 1.5 DAP ovaries, bGlcY
treatments had no effects on the size of pre-fertilized
ovaries (Fig. 6A), but significantly impact the length and
width of pre-fertilized ovules within the cultured ovaries as
found for the total analysis of ;600 ovules of each group
(Fig. 6B). When the concentration of bGlcY increased,
the size of the ovules decreased (Fig. 6B). The sizes of
post-fertilized ovaries were not remarkably affected by the
treatments with bGlcY (Fig. 6C), but the sizes of postfertilized ovules were decreased by the treatment with
high concentrations (50 and 100 lM) of bGlcY after 4 d
of culture (Fig. 6D). The control aGalY reagent did not
influence the growth of ovaries and ovules.
To clarify whether the growth inhibition of the ovules by
bGlcY treatment is due to arrested cell division or cell
expansion, the cell number and cell size were examined
in semi-thin sections of the ovules. Figure 7 shows that
the control and bGlcY-treated ovules have a similar cell
number but different cell size, indicating that bGlcY
2068 Qin and Zhao
Fig. 5. Growth and development of zygotes at 3.5 DAP in in vitro culture. (A–F) Zygotes cultured in the medium without bGlcY. (A) An isolated
spherical zygote. (B) An isolated elongated zygote. (C) An elongated zygote undergoing first asymmetrical division. (D) A spherical zygote undergoing
first symmetrical division. (E) A multicellular proembryo derived from a zygote. (F) CW fluorescence image of the proembryo in (E) showing the cell
wall. (G–I) Zygotes cultured in medium supplemented with 50 lM bGlcY. (G) Symmetrical division of a zygote. (H) Budded division of a zygote. (I)
Asymmetrical division of a zygote. (J–M) Zygotes cultured in medium supplemented with 80 lM bGlcY. (J) Symmetrical division of a zygote. (K) A
multicellular structure without an embryo body and suspensor derived from a symmetric divided zygote. (L) Asymmetrical division of a zygote. (M) A
proembryo derived from an asymmetric divided zygote with a typical structure containing an embryo body and suspensor. (N–P) Zygotes cultured in the
medium supplemented with 100 lM bGlcY. (N) Symmetric division of a zygote. (O) Asymmetric division of a zygote. (P) A four-cell proembryo
derived from an asymmetric divivided zygote. Arrowheads show the division planes of the zygotes. Bar=10 lm.
impacted cell expansion but not cell division during ovule
growth. Both integument cells and nucellar cells in the
ovule treated with 100 lM bGlcY (Fig. 7B) are apparently
smaller than those of the control without treatment (Fig.
7A). aGalY is an analogue of bGlcY but is incapable of
binding AGPs. The cell number and cell size of the ovules
treated with 100 lM aGalY (Fig. 7C) are essentially the
same as those of the non-treatment control (Fig. 7A). The
results suggested that AGPs play a crucial role in cell
expansion during ovule development. Figure 7D shows
the statistical analysis of the area of nucellar cells in the
sections. It is interesting that the morphologies of cells in
Egg cell fertilization and zygote division in N. tabacum L.
2069
Table 1. Effects of bGlcY on zygote division during in vitro zygote culture
Concentration of
bGlcY (lM)
No. of cultured
zygotes
First zygote division
No. of
divided zygotes
0
10
50
80
100
72
57
48
45
37
63
48
39
35
29
Frequency of divided
zygotes (%)
87.5
84.2
81.2
77.8
78.4
Modes of division
Symmetrical
division (%)
Asymmetric
division (%)
Budded
division (%)
5.6
8.8
27.1
37.8
37.8
79.2
75.4
31.2
40.0
37.8
2.7
0
22.9
0
2.8
Fig. 6. Impacts of bGlcY and aGalY (a Yariv reagent which does not bind AGPs) on the growth of ovaries and ovules at 1.5 DAP (A, B) and 3 DAP
(C, D) in in vitro ovary culture. The sizes were measured after 4 d of culture. Values are means (6SE) from 5–6 independent experiments with 25
(A, C) or 600 (B, D) replicates, and the P-values were calculated according to the untreated series as controls. No asterisk indicates the means of no
difference (P >0.05). One asterisk indicates the means of a distinct difference (0.01<P<0.05). Two asterisks indicate the means of a very distinct
difference (P <0.01).
the bGlcY-treated samples have a rather crushed appearance (altered shape), rather than just appearing smaller (Fig.
7B). One of the possible reasons is that blocking AGP
activity through bGlcY treatment has perturbed the structural integrity of the walls of nucellar cells. Thus it also
implied that AGPs function in determining cell shape.
After 4 d of culture of 1.5 DAP ovaries in normal culture
conditions, both egg cells and central cells were fertilized and
developed into elongated zygotes (Fig. 8A) and multicellular
endosperms, respectively. However, treatment with 100 lM
bGlcY resulted in failure of fertilization and zygote formation. The egg cells in this treatment condition displayed the
characteristics of an unfertilized egg cell, i.e. a large vacuole
located at the micropylar end of the cells and the nucleus and
cytoplasm at the chalazal end (Fig. 8B, C). The unfertilized
egg cells within the ovaries treated with 100 lM bGlcY
showed signs of cell degeneration such as a degenerated
nucleus, disorganized nuclear membrane, and diffused
nuclear substances evidenced by ultrathin sections (Fig.
8C), which resemble natural unfertilized egg cells in the
tobacco plant during sexual reproduction. Although fertilization of egg cells was severely hampered, the fertilization
and subsequent division of central cells could still occur
under treatment with bGlcY (Fig. 8B).
2070 Qin and Zhao
Fig. 7. Impacts of bGlcY and aGalY on cell division and cell expansion in ovules in in vitro ovary culture. Samples for sectioning were collected after
4 d of ovary culture. (A) Semi-thin section of an ovule in the control test. (B) Semi-thin section of an ovule treated with 100 lM bGlcY. (C) Semi-thin
section of an ovule with control treatment with 100 lM aGalY. (D) Measurement of the cell areas in sections of ovules. Treatment with 100 lM bGlcY
decreased the sizes of tobacco nucellar cells. ES, embryo sac; I, integument; NC, nucellar cells. Bar=20 lm.
After 4 d of culture in normal conditions, zygotes within
3 DAP ovaries could divide to form globular proembryos at
different stages. Table 2 shows the effects of bGlcY on
early embrygenesis in in vitro ovary culture. Along with an
increase in the concentration of bGlcY, the formation of
two-celled proembryos increased and the formation of
3–16-cell proembryos was not significantly affected, while
the number of proembryos containing >17 cells decreased
dramatically. In the control tests without bGlcY or with
100 lM aGalY in the medium, two-celled proembryos
comprised only 3.57% and 8.69%, respectively, of the total
proembryos formed, while proembryos consisting of >17
cells accounted for 39.29% of total proembryos. In contrast,
the number of two-celled proembryos increased while the
number of proembryos with >17 cells decreased dramatically under bGlcY treatments (Table 2). These results
suggest that bGlcY arrested early embryo development.
Resembling its effects on zygotic division during zygote
culture, bGlcY also impacts the pattern of zygotic division
occurring within cultured ovaries. Treatments of the ovary
cultures with bGlcY substantially increased symmetrical
division, but decreased asymmetrical division of zygotes
(Table 3). Morphological observations revealed that typical
asymmetrical division of zygotes in the control cultures
resulted in a two-celled proembryo with a small apical cell
and a large basal cell (Fig. 8D, E), while symmetrical
division of zygotes resulting from bGlcY treatments led
to the formation of proembryos with two indistinctive cells
(Fig. 8F, G). Some abnormal symmetrical two-celled
proembryos could continue symmetrical division to form
four-celled proembryos without a typical embryo proper
and suspensor (Fig. 8H, I). These four-celled proembryos
contain four almost identical cells as revealed by semi-thin
sections (Fig. 8I). Some two-celled proembryos resulting
from asymmetrical zygote division further developed into
four-celled proembryos (Fig. 8J) and globular proembryos
consisting of a typical embryo proper and suspensor
(Fig. 8K).
Discussion
Developmental regulation of AGP expression during
plant sexual reproduction
Dramatic changes of AGPs during ovule growth and
development were detected by western blotting in this
study. In particular, the expression of AGPs recognized by
LM2 is apparently developmentally regulated in ovules before and after fertilization (Fig. 1B, C). High accumulation
of AGPs in mature ovules before fertilization suggests
that AGPs may play an important role in directing pollen
tube growth into ovules. Directional cues originating from
ovules are believed to play an important role in dramatic
bending of the pollen tube tip to gain access to the ovules
Egg cell fertilization and zygote division in N. tabacum L.
2071
Fig. 8. Impacts of bGlcY and aGalY on the development of egg cells and zygotes during in vitro ovary culture. Samples for sectioning and proembryo
isolation were collected after 4 d culture of ovaries at 1.5 DAP (A–D) and 3 DAP (E–K). (A) An ultrathin section showing an elongated zygote in ovary
culture without bGlcY as a control test. (B) A section of an ovule treated with 100 lM bGlcY. Note that the egg cell could not fertilize and form a zygote,
but still retained the characteristics of an egg cell with a large vacuole at the micropylar end, and cytoplasm and nucleus at the chalazal end. (C) Ultrathin
section showing degeneration of an unfertilized egg cell with a disorganized nuclear membrane after 4 d culture of the ovary in medium with 100 lM
bGlcY. (D) An isolated normal asymmetric two-celled proembryo from an ovary cultured in medium without bGlcY. (E) Semi-thin section showing
a two-celled proembryo consisting of a small apical cell and a large basal cell after ovary culture in medium supplemented with 100 lM aGalY. (F) An
isolated abnormal symmetric two-celled proembryo from an ovary cultured in medium with 100 lM bGlcY. (G) Semi-thin section showing an abnormal
two-celled proembryo derived from symmetric division of a zygote. (H) An isolated symmetric 4-cell proembryo from an ovary cultured in medium with
100 lM bGlcY. (I) Semi-thin section showing an abnormal four-cell proembryo consisting of four nearly identical cells in an ovary cultured in medium
with 100 lM bGlcY. (J) An isolated normal asymmetric four-cell proembryo from an ovary cultured in medium without bGlcY. (K) An isolated globular
proembryo consisting of an embryo proper and suspensor from an ovary cultured in medium without bGlcY. AC, apical cell; BC, basal cell; EC, egg cell;
EN, egg cell nucleus; PS, persistent synergid; ES, endosperm; ZN, zygotic nucleus. Arrowheads show the zygotic first division planes and arrows show
the planes of subsequent divisions. Bar=10 lm.
where the egg cells and central cells are located. Coimbra
and Salema (1997) reported that the expression of MAC207
and JIM8 epitopes seems to be closely connected with the
pathway of pollen tube growth in Amaranthus hypochondriacus L. ovules. They proposed that the distribution
of AGPs in ovules might provide signal guidance for
pollen tube growth in order to reach the egg cells and
central cells.
Besides the change in AGPs recognized by LM2 during
fertilization, significant alteration of this type of AGP
during embryo development was also detected. A low level
of LM2 epitope is present in ovules in which proembryos
are at early (5 DAP) and middle (6 DAP) globular stages
(Fig. 1B). However, AGPs recognized by LM2 increased in
7 DAP ovules in which globular embryos start differentiation, and the amount of AGPs remained constant in 8
and 9 DAP ovules where embryogenesis is at the stage of
heart-shaped embryo and torpedo-shaped embryo, respectively. These results indicate that the AGPs recognized
by LM2 might be important for cell differentiation during
embryogenesis. In contrast, no dramatic change was detected in the expression of the JIM13 epitope during ovule
2072 Qin and Zhao
Table 2. Effects of bGlcY on early embryo development in in vitro-cultured ovaries
No. of isolated
proembryo cells
Control
(no bGlcY) (%)
10 lM
bGlcY (%)
50 lM
bGlcY (%)
100 lM
bGlcY (%)
100 lM
aGalY (%)
2
3–4
5–8
9–16
>17
No. of isolated proembryos
3.57
17.86
16.07
23.21
39.29
56
11.36
22.74
18.18
20.45
27.27
44
18.64
22.04
20.34
18.64
20.34
59
23.08
26.15
18.46
21.54
10.77
65
8.69
19.57
15.22
21.74
34.78
46
Table 3. Effects of bGlcY on the first division of zygotes in in vitro-cultured ovaries
Treatments
Control
10 lM bGlcY
50 lM bGlcY
100 lM bGlcY
100 lM aGalY
No. of isolated proembryos
56
44
59
65
46
No. of two-celled embryos
2
5
11
15
4
development, which implies that different sets of AGPs
are involved in particular aspects of plant growth and
development.
Roles of AGPs in fertilization
Mb JIM8 was generated against the carbohydrate of AGPs
on plasma membranes of carrot suspension-cultured cells,
and thus recognizes plasma membrane AGPs. In oilseed
rape ovules, a polar distribution of the JIM8 epitope was
also observed. The JIM8 epitope is present in the egg cell
and both synergids, but is absent in the central cell (Pennell
et al., 1991). Isolated sperm cells of Brassica campestris
and Lilium longiflorum were also shown to express JIM8
and JIM13 epitopes (Southworth and Kwiatkowski, 1996).
Using immunogold electron micrography, Pennell et al.
(1991) has shown that the JIM8 epitope is on the outer
plasma membrane of sperm cells in oilseed rape pollen
(Pennell et al., 1991). The presence of AGPs on the cell
surface of both egg cells and sperm cells implies their roles
in egg–sperm cell recognition during fertilization. Using
immunofluorescent labelling, a strong fluorescence signal
was detected revealing intensive expression of AGPs recognized by JIM13 at the micropylar end of the embryo
sac (Fig. 2A). In contrast, the central cell displayed weak
fluorescence. This observation was further confirmed by
immunofluorescence labelling of isolated egg cells and
central cells (Fig. 2B, D). Differential expression of JIM13labelled AGPs in egg cells and central cells suggests that
AGPs might be required for fertilization of egg cells but
are not essential for fertilization of central cells. This notion
is also well supported by our observation that fertilization
of egg cells but not central cells in cultured ovaries was
perturbed by bGlcY treatment (Fig. 8B, C).
Modes of division
Symmetric division
Asymmetric division
No.
Frequency (%)
No.
Frequency (%)
0
1
5
7
0
0
20
45.45
46.67
0
2
4
6
8
4
100
80
54.55
53.33
100
One interesting observation from our study is the
dramatic change in AGP abundance in the plasma membrane of pre- and post-fertilized egg cells. A uniform and
strong fluorescent ring and intensive gold particles in the
plasma membrane of pre-fertilized egg cells (Figs 2B, 3A)
reveal high accumulation of AGPs on the surface of
unfertilized egg cells. In contrast, the AGP content is dramatically decreased in fertilized egg cells (Figs 2E, 3C).
The change of AGPs in egg cells before and after fertilization may play a crucial role in the expansion of the
egg cell and zygote. AGP proteins encoded by SOS5 have
been shown to be required for normal cell expansion (Shi
et al., 2003). A high level of AGPs on the surface of
unfertilized egg cells might be important to restrict cell
expansion before fertilization. However, after fertilization,
the most immediate change is elongation of zygotes,
which is required for the subsequent asymmetric zygote
division. A decrease of AGPs on the surface of fertilized
egg cells might facilitate cell expansion during the zygote
elongation process. On the other hand, zygote elongation
is accompanied by the migration of the nucleus from the
central part of the cell to the region near to the chalaza.
Polar distribution of AGPs in the elongated zygote (Fig.
2F) indicates that AGPs might also be involved in
establishing and stabilizing zygotic polarity, which is
the first step determining cell fate during early embryo
development (Vroemen et al., 1999).
Effects of AGPs on zygotic division
bGlcY has been widely used to perturb AGP functions in
living cells because of its capacity to bind to and thus
inactivate AGPs (Knox, 1997). Much attention has been
Egg cell fertilization and zygote division in N. tabacum L.
paid to the roles of AGPs in embryogenesis. Thompson and
Knox (1998) employed an in vitro cell suspension culture
system in Daucus carota and studied the effect of bGlcY
on somatic embryogenesis. Chapman (2000) showed the
bGlcY concentration-dependent inhibition of somatic embryogenesis in Cichorium. Because egg cells and zygotes
are deeply embedded inside the ovule and ovary, their
inaccessibility has hindered our understanding of the role
of AGPs in zygotic division in higher plants. In this
study, in vitro zygote and ovary culture systems was used
to facilitate the analysis of AGP functions in zygotic
division by using bGlcY.
These studies indicate that bGlcY treatment greatly
impacted the mode of zygotic division rather than the
division frequency. bGlcY treatment increased atypical
symmetric division of zygotes and decreased normal
asymmetric zygote division. These results suggest that
AGPs play crucial roles in establishing the polarity of the
zygote and the subsequent asymmetric division in planta,
which is consistent with our observation that AGPs are
asymmetrically distributed in the elongated zygote (Fig.
2F). The cytoskeleton is well known to be involved in
establishing and stabilizing the polar axis and in determining the fate of cells in early proembryo development
(Vroemen et al., 1999). Disruption of AGPs by bGlcY
might perturb microtubule alignment and lead to reorganization of the cytoskeleton during zygote division.
Many challenging questions regarding the functions of
AGPs in plant sexual reproduction remain to be answered.
Further studies are expected to elucidate the precise
functions of AGPs in fertilization, embryogenesis, and
embryonic cell differentiation.
Acknowledgements
The authors thank Dr JP Knox (Centre for Plant Sciences, University
of Leeds, UK) for the generous gifts of the antibodies, and Assistant
Professor Huazhong Shi (Department of Chemistry and Biochemistry, Texas Tech University, USA) for his editorial assistance. This
project was supported by the National Natural Science Foundation of
China (30370089, 30521004) and by the Program for Changjiang
Scholars and Innovative Research Team (PCSIRT) at University in
China.
References
Chapman A, Blervacq AS, Vasseur J, Hilbert JL. 2000.
Arabinogalactan-proteins in Cichorium somatic embryogenesis:
effect of b-glucosyl Yariv reagent and epitope localization during
embryo development. Planta 211, 305–314.
Coimbra S, Salema R. 1997. Immulocalization of arabinogalactan
proteins in Amaranthus hypochondriacus L. ovules. Protoplasma
199, 75–82.
Egertsdotter U, Arnold SV. 1995. Importance of arabinogalactan
proteins for the development of somatic embryos of Norway
spruce (Picea abies). Physiologia Plantarum 93, 334–345.
2073
Faure JE, Digonnet C, Dumas C. 1994. An in vitro system
for adhesion and fusion of maize gametes. Science 263, 1598–
1600.
Fincher GB, Stone BA, Clarke AE. 1983. Arabinogalactanproteins: structure, biosynthesis and function. Annual Review of
Plant Physiology 34, 47–70.
Knox JP. 1996. Arabinogalactan-proteins: developmentally regulated proteoglycans on the plant cell surface. In: Smallwood M,
Knox JP, Bowles DJ, eds. Membranes: specialized functions in
plants. Oxford: BIOS Scientific Publishers, 93–102.
Knox JP. 1997. The use of antibodies to study the architecture and
developmental regulation of plant cell walls. International Review
of Cytology 171, 79–120.
Knox RB, Southworth D, Sing MB. 1988. Sperm cell determinants
and control of fertilization in plants. In: Chapman GP, Ainsworth
CC, Chatham CJ, eds. Eukaryote cell recognition: concepts
and model systems. Cambridge: Cambridge University Press,
175–193.
Kreuger M, van Holst GJ. 1993. Arabinogalactan-proteins as
essential in somatic embryogenesis of Daucus carota L. Planta
189, 243–248.
Li ST, Yang HY. 2000. Gene transfer into isolated and cultured
tobacco zygotes by a specially designed device for electroporation.
Plant Cell Reproduction 19, 1184–1187.
Majewska-Sawka A, Nothnagel EA. 2000. The multiple roles of
arabinogalactan proteins in plant development. Plant Physiology
122, 3–9.
Nothnagel EA. 1997. Proteoglycans and related components in plant
cells. International Review of Plant Cytology 174, 195–291.
Pennell RI, Janniche L, Kjellbom P, Scofield GN, Peart JM,
Roberts K. 1991. Developmental regulation of a plasma membrane arabinogalactan protein epitope in oilseed rape flowers.
The Plant Cell 3, 1317–1326.
Pennell RI, Roberts K. 1990. Sexual development in pea is
presaged by altered expression of arabinogalactan protein. Nature
34, 547–549
Russell SD. 1985. Preferential fertilization in Plumbago: ultrastructural evidence for gamete-level recognition in an angiosperm.
Proceedings of the National Academy of Sciences, USA 82,
6129–6132.
Shi HZ, Kim Y, Guo Y, Stevenson B, Zhu JK. 2003. The
Arabidopsis SOS5 locus encodes a putative cell surface adhesion
protein and is required for normal cell expansion. The Plant Cell
15, 19–32.
Showalter AM. 2001. Arabinogalactan-proteins: structure, expression and function. Cellular and Molecular Life Science 58,
1399–1417.
Southworth D, Kwiatkowski S. 1996. Arabinogalactan proteins at
the cell surface of Brassica sperm and Lilium sperm and generative
cells. Sexual Plant Reproduction 9, 269–272.
Stacey NJ, Roberts K, Knox JP. 1990. Patterns of expression of the
JIM4 arabinogalactan-protein epitope in cell cultures and during
somatic embryogenesis in Daucus carota L. Planta 180, 285–292.
Tian HQ, Russell SD. 1997. Micromanipulation of male and female
gametes of Nicotiana tabacum: I. Isolation of gametes. Plant Cell
Reproduction 16, 555–560.
Thompson HJM, Knox JP. 1998. Stage-specific responses of
embryogenic carrot cell suspension cultures to arabinogalactan
protein-binding b-glucosyl Yariv reagent. Planta 205, 32–38.
Vroemen C, de Vries S, Quatrano R. 1999. Signalling in plant
embryos during the establishment of the polar axis. Cell and
Developmental Biology 10, 157–164.
Wittink FRK, Knuiman B, Derksen J, Capkova V, Twell D,
Schrauwen JAM, Wullems GJ. 2000. The pollen-specific gene
Htp303 encodes a 69 kDa glycoprotein associated with the
2074 Qin and Zhao
vegetative membranes and the cell wall. Sexual Plant Reproduction 12, 276–284.
Wu H, de Graaf B, Mariani C, Cheung AY. 2001. Hydroxyprolinerich glycoproteins in plant reproductive tissues: structure, functions and regulation. Cellular and Molecular Life Science 58,
1418–1429.
Zhao J, Mollet JG, Lord EM. 2004. Lily (Lilium longiflorum
L.) pollen protoplast adhesion is increased in the presence
of the peptide SCA. Sexual Plant Reproduction 16, 227–233.
Zhu JK, Bressan RA, Hasegawa PM. 1993. Loss of arabinogalactan-proteins from the plasma membrane of NaCl-adapted
tobacco cells. Planta 190, 221–226.