Content uploaded by Nina Gabarayeva
Author content
All content in this area was uploaded by Nina Gabarayeva on Jan 21, 2019
Content may be subject to copyright.
Sporoderm development in Trevesia burckii (Araliaceae). I. Tetrad period:
Further evidence for the participation of self-assembly processes☆
Nina Gabarayeva
a,
⁎, Valentina Grigorjeva
a
, John R. Rowley
b
, Alan R. Hemsley
c
a
Komarov Botanical Institute, Popov st., 2, 197376, St.-Petersburg, Russia
b
Botanical Institute, Stockholm University, Stockholm, SE-106 91, Sweden
c
Department of Earth Sciences, Cardiff University, Park Place, Cardiff CF10 3YE, Wales, UK
abstractarticle info
Article history:
Received 22 February 2008
Received in revised form 28 November 2008
Accepted 2 December 2008
Available online 7 December 2008
Keywords:
microstructure
pattern formation
self-assembly
colloids
tensegrity
ECM-glycocalyx
pollen wall development
callose envelope
Trevesia
The developmental events in the periplasmic space and cytoplasm of microspores, and in the tapetum of
Trevesia burckii during the tetrad period have been traced in detail during microspore ontogeny from the
sporogenous cell stage to the late tetrad stage and the initiation of the foot layer. The data obtained provide
support to two of our previously proposed hypotheses: (1) the glycocalyx (scaffolding primexine matrix) is a
colloidal system; (2) the involvement of processes of self-assembly of a number of colloidal micellar systems
to the exine development. The main structures of the reticulate ectexine up to the establishment of the foot
layer are columellae and tectum which evidently form on a base of spherical and cylindrical transitive micelle
mesophases after sporopollenin accumulation. The importance of the callose envelope surrounding
microspores for exine development is discussed in the light of recent findings. Two possible pathways of
pattern determination are possible. One suggests the role of the plasma membrane in pattern imprinting and
the corresponding necessity of the transfer of 2-D information to 3-D. Our current supposition is that the
other, self-assembly physical phenomena, cellular tensegrity, also participate in the process of the
establishment of a wide-spread reticulate exine pattern which appears as the result of an interplay of the
microspore cytoskeletal prestress and the resistance from the ECM (exocellular matrix=glycocalyx)
adhesive sites. More and more information is appearing that provides evidence for the importance of
mechanical forces and physico-chemical regularities in the living world.
© 2008 Elsevier B.V. All rights reserved.
1. Introduction
In spite of the fact that a great many recent studies have been
undertaken on spore and pollen wall development, our knowledge of
exine patterning is still obscure. However, if we ever hope to unravel
this enigma, we will not manage this without developmental insights
which provide a route to the understanding of the processes involved
in generating characteristic patterns of pollen or spore wall structure
and ornamentation —the patterning, originating during development
in response to the subtle variations of the conditions within the
plasma membrane glycocalyx (Blackmore et al., 2007).
The question arises as to where such a key exists? The often
species-specific nature of exine patterns suggests a genomic key. But
as more facts accumulate, the more suspicions appear that something
interferes with the precise and straightforward work of the genome.
This “something”is, in the opinion of a number of authors, the inter-
ference of self-assembly processes which are capable of distorting the
regular work of the genome, making the results unpredictable
(Heslop-Harrison, 1972; Gerasimova-Navashina, 1973; Sheldon and
Dickinson, 1983; Gabarayeva, 1990; van Uffelen, 1991; Hemsley et al.,
1992; Gabarayeva, 1993; Southworth and Jernstedt, 1995). It is a
matter of course that the input of the genome is large: without doubt,
the genome determines the exact chemical composition of all the
substances in the periplasmic space and their concentration, the rest
of constructive process is picked up by physico-chemical self-
assembly (Gabarayeva and Hemsley, 2006). This is the point where
possible distortion of initial genomic instruction could happen: the
non-linear character of self-assembly. The suspicions come also
from other sources: the striking similarity of lifeless patterns
(e.g. technology organometallic covers —see Fig. 1 in Hemsley and
Gabarayeva, 2007) and exine microarchitecture is really confusing.
The meaningful title of Hemsley et al's. paper (2004) includes the
phrase “soft and sticky development”which points to the participa-
tion of “soft matter”in exine development. The term “soft matter”
denotes materials that are easily deformed by external stress, and
encompasses liquid crystals, polymers, surfactants and colloids —
particles dispersed within another medium (van Blaaderen, 2006).
Review of Palaeobotany and Palynology 156 (2009) 211–232
☆Nature need not cram genetic code with detail, predetermined by self-assembly
(adapted from “The Fractal Geometry of Nature”by B.B. Mandelbrot (1982).
⁎Corresponding author.
E-mail address: 1906ng@mail.ru (N. Gabarayeva).
0034-6667/$ –see front matter © 20 08 Elsevier B.V. All rights reserved.
doi:10.1016/j.revpalbo.2008.12.001
Contents lists available at ScienceDirect
Review of Palaeobotany and Palynology
journal homepage: www.elsevier.com/locate/revpalbo
Their basic constituents have sizes of between several nanometres and
several micrometres, and have the potential to self-organize, forming
beautiful 3-D structures —complex colloids on scales up to a
micrometre. Interestingly, the thickness of the exine in most species
is about 1 mm. These structures, obtained experimentally by a number
of authors (see review of van Blaaderen, 2006), are whimsical: “ice-
cream cones”, the structures, indistinguishable from polyporate
pollen grains, have “dome-like”and “raspberry-like”structures. The
innovation is that these structures are being formed with a second
stage of self-organization. The particles are first formed as soft-matter
scales, and then built to a far more intricate structure, allowing
unprecedented control over the 3-D organization. Raspberry-like
mimics were obtained from polystyrene latex dispersions in experi-
ments modeling exine microarchitecture (Griffiths and Hemsley,
2001). The most important conclusion is, however, that complex
shape is no prerequisite for complex interaction. Another important
feature of soft-matter components is that these structures can be
recovered in the process of drying of emulsion droplets. It is
meaningful that the early surmise of Rowley (Rowley and Rowley,
1998) was that late exine development (especially concerning pollen
wall ornamentation) proceeds very quickly in desiccating conditions
before pollen dispersion. Some soft-matter structures were fabricated
by deposition of silica on liquid crystals formed by surfactants (Zoldesi
and Imhof, cited by van Blaaderen, 2006). It is significant that Tryon
and Lugardon (1978) revealed silicon and calcium depositions in the
megaspores of Selaginella by X-ray spectrography and discussed
mineral deposition in relation to the occurrence of this material in
other spores and the general role of these elements in other plant
structures.
In our previous papers we suggested a hypothesis for the
explanation of several stages of spore/microspore wall development
on the basis of physical–chemical self-assembly processes unfolding in
colloidal micelle systems in the periplasmic space (Gabarayeva and
Hemsley, 2006; Hemsley and Gabarayeva, 2007). At other stages direct
genomic control prevails (Gabarayeva and Hemsley, 2006 —see Fig. 6
in that paper). The scheme of an elementary micelle system, with their
reversable mesophases, is summarized in Fig. 1. Our objective in this
study was to show in detail the exine development in Trevesia burckii
up to the late tetrad stage, and to confirm this hypothesis, while in this
paper we concentrate on the physico-chemical underlying cause. We
also suggest the involvement of some other self-assembling mechan-
isms (tensegrity, prestress) for exine pattern establishment.
The Trevesia species have been under consideration by several
authors (Xu Ting-Yu, 1982; Grushvitsky et al., 1984; Shang Chih-Bei
and Callen, 1988).
The abbreviations used in this paper are as follows:
CSK cytoskeleton
3-D three-dimensional
ECM extra (= exocellular) matrix
MF microfilament
MT microtubule
SAPs sporopollenin acceptor particles
SP sporopollenin
Fig. 1. Transitive stages of micelles and micelle aggregations (adapted from Hamley (2000) and Fridrichsberg (1995) which appear in solution as the concentration of surfactant
increases. Sketches b, c, d, e show sequential micelle mesophases formed in a water-based medium (so-called direct micelles). Sketches b', c', d', e show corresponding micelle
mesophases in a lipid-based medium (so-called reverse micelles). Note that stage e is the same for both sequences. a —true solution of surfactant molecules, with their hydrophilic
head and hydrophobic tails. b —spherical micelle, with hydrophilic heads outside and hydrophobic tails hidden inside, formed at critical micelle concentration. c —cylindrical micelle
formed at higher surfactant concentration. d —hexagonally packed cylindrical micelles spontaneously form a layer —“middle”mesophase. e —lamellar “neat”micelles in parallel
arrangement, formed by bilayers of “middle”mesophase. Arrowheads show typical gaps between bilayers. b' —reverse spherical micelle, with hydrophobic tails sticking out and
hydrophilic heads inside. c' —reverse cylindrical micelle. d' —layer of reverse hexagonally packed cylindrical micelles (reverse “middle”mesophase). Note that if a water-based
medium changes to a lipid-based one, “direct”micelles undergo inversion. Note also the opposite course of events (see arrows).
212 N. Gabarayeva et al. / Review of Palaeobotany and Palynology 156 (2009) 211–232
2. Materials and methods
Flower buds of Trevesia burckii Boerl. (Araliaceae) were collected
from plants (inventory number 259740) in the greenhouses of the
Komarov Botanical Institute, St.-Petersburg, over a period of five years.
Fragments of stamens were fixed in 3% glutaraldehyde and 2.5%
sucrose in 0.1 M cacodolate buffer (pH 7.3, 20 °C, 24 h), with the
addition of 1% of lanthanum nitrate. The material was post-fixed with
2% osmium tetroxide (pH 8.0, 20 °C, 2 h). After acetone dehydration
the samples were embedded in a mixture of Epon and Araldite.
Ultrathin sections were cut with a diamond knife and stained with a
saturated solution of uranyl acetate in ethanol and 0.2% lead citrate.
Sections were examined with a Hitachi H-600 TEM.
3. Results
3.1. Pretetrad period
Sporogenous cells with angular outlines fill the loculus of the
anther and are surrounded with tapetal cells (Plate I, 1). The latter
have very dense cytoplasm and show darker contrast than sporogen-
ous cells. Large nuclei with one or two nucleoli occupy the central part
in the cytoplasm of the sporogenous cells where numerous small
vacuoles are observed. Cytomictic channels connect the cytoplasm of
the adjacent tapetal cells (Plate I, 1, arrows). One is shown in Plate I,2
(arrow). Later, at the stage of microspore mother cells covered with a
thin callose envelope (Plate I, 3), some tapetal cells develop
outgrowths and penetrate between adjacent microspore mother
cells, whereas other tapetal cells remain in position. The tapetal
cells are in a state of hyperactivity and contain huge osmiophilic
plastids and ER dilations. After disintegration of the nuclear envelope
meiosis occurs. A metaphase is shown in Plate I, 4. On meiosis
completion, the tetrad period of microspore development begins.
3.2. Tetrad period
3.2.1. Very young tetrad stage
The onset of this post-meiotic stage is marked by the appearance of
a thick callose envelope around each young microspore and the tetrad
as a whole (Plate I, 5). The profile of the plasma membrane is rather
even and lacks any signs of the cell surface coating (glycocalyx) at this
point in ontogeny (Plate I, 6). This changes when, some time later, the
profile of the plasma membrane becomes wavy and dark-contrasted
globules appear in the periplasmic space on the surface of plasma
membrane (Plate II, 1). The surface of tapetal cells also is covered
with the same material (Plate II, 2), a factor critical in the develop-
mental sequence. At higher magnification some of the osmiophilic
globules show small ‘cogs’on their surface (Plate II, 3). Endocytotic
channels and vacuoles, with engulfed portions of the osmiophilic
material covering the plasma membrane, are observed in the cyto-
plasm at this stage (Plate II, 4). The radial extension of the periplasmic
space at sites of the plasma membrane invaginations is about 500 nm.
This is indicative of very active behaviour of the plasma membrane in
the process of endocytosis. The invasive tapetal cells also bear
osmiophilic droplets on their surface (Plate II, 5). Somewhat later a
diversity of structures are seen in the periplasmic space: globules of
different size; rod-like or rather tiny tubule-like structures (evidently
attached at the base to the plasma membrane), globules in asso-
ciation with tiny tubules which have a form resembling tadpoles; and
even structures that appear like some kind of laminated fern leaves
(Plate II, 6).
All these structures disappear by the next stage, (probably being
engulfed in the process of endocytosis) and relinquish their place in
the periplasmic space to the real glycocalyx (i.e. these stages have
been transitory, preparatory, pre-exine prior to initiation of the next
ontogenetic event —exine establishment).
3.2.2. Young tetrad stage
A survey of a part of a tetrad is shown in Plate III, 1. The plasma
membrane of the microspores at this stage has a rather even profile
and is covered with a thin layer of the glycocalyx, its radial extent is
about 100 nm. The glycocalyx layer consists mainly of rounded units
with a dark-contrasted core and light-contrasted halo, in addition to
rare radially oriented rods (Plate III,2–3). The glycocalyx material is
delivered into the periplasmic space by Golgi vesicles. Indeed,
microspore cytoplasm is full of Golgi vesicles and also contains coated
vesicles (Plate III, 3). At the next ontogenetic step the plasma
membrane is again strongly invaginated, and the radius of the peri-
plasmic space is up to 250 nm (Plate III, 4). Two generations of the
glycocalyx are seen within the periplasmic space: (1) a fine network-
like pattern, with the addition of membrane-like structures, and (2) a
dark-contrasted layer covering the plasma membrane which consists
of rounded units with an osmiophilic core and a less contrasted halo
(Plate III, 4). Here and there rounded units are seen in radial piles (the
central arrow in Plate III, 4) or show short branches (the left arrow in
Plate III, 4). The cytoplasm is overcrowded with ribosomes and
contains dilated cysternae of ER, that are mainly RER with dark
contrasted contents (Plate III, 4). More distinctly, clusters of rounded
units, and somewhat later their radially oriented associations, are
observed (Plate III, 5). These are shown in more detail in Plate III,6.
Dilated ER, Golgi stacks and Golgi vesicles are seen in the peripheral
cytoplasm of the microspore (Plate III, 5).
3.2.3. The completion of the young tetrad microspore stage
The microspore plasma membrane profile becomes periodically
invaginated, with increased radius of the periplasmic space in
invaginations of up to 500 nm (Plate IV, 1, 2). This stage is a transitory
one before the next middle tetrad stage with its regularly invaginated
plasma membrane. Images in Plate IV, 3 and 4 most probably show the
incomplete process of pre-patterning of the future reticulate exine
pattern. Invagination sites, as in the previous stage, contain two
successively formed glycocalyx generations: (1) a fine network,
situated distally in the periplasmic space (closer to the callose —
marked by stars), and (2) more proximal areas of darker material
which are represented by radially-oriented, extended units attached
to the plasma membrane (Plate IV, 3, 4, marked by asterisks). These
extended rod-like units are probably in the process of construction of
the rounded units with a dark core and light halo (the left invagination
area in Plate IV, 3). The rounded units at this stage are of smaller
diameter than those at previous stage (Plate III, 5) and are a new
feature of the periplasmic space. It is probable that the previous
generation of rounded units was engulfed and seen in the cytoplasm
as a cluster (Plate IV, 4). Close observation also reveals rod-like units
on the top of evaginated parts of the plasma membrane. This, and
aspects of the next development stage (Plate V), lead to the conclusion
that large invaginated areas such as those marked by asterisks in
Plate IV, 3 and 4, will be later raised in development. The microspore
cytoplasm is packed with active Golgi stacks and their vesicles and
contains cysternae of RER and small vacuoles.
3.2.4. Middle tetrad stage
Plate V, 1 shows the border of the anther with a row of tapetal cells
and an adjacent tetrad. The surface of the tetrad microspore is deeply
and regularly invaginated. At higher magnification (Plate V,2)
procolumellae are distinctly seen on the tops of the evaginated sites.
The maximum radius of the glycocalyx in areas of invaginations is
about 750 nm. In the periplasmic space both generations of the
glycocalyx, (observed at the previous stage in Plate IV,3–4), are seen
clearly and marked correspondingly by stars and asterisks, but the
second generation (marked by asterisks) has lost its connection with
the plasma membrane inside invaginated areas and will grow no more
in ontogeny, whereas on the tops of evaginations this latter glycocalyx
generation experiences intensive growth from the side of plasma
213N. Gabarayeva et al. / Review of Palaeobotany and Palynology 156 (2009) 211–232
membrane resulting in the appearance of procolumellae. In detail,
procolumellae at this stage are bundles of radially-oriented extended
tubular units, with a dark core and light halo, the free ends of which
are seen at high magnification in Plate V, 3. Above each bundle
(procolumella) an electron-transparent zone is observed. Its transpar-
ency makes the appearance of free ends of the radial units especially
distinctive. The microspore cytoplasm is full of ribosomes and
contains small vacuoles, tubular ER and low-differentiated mitochon-
dria, whereas the cytoplasm of tapetal cells contains stacked RER,
plastids with osmiophilic inclusions, large vacuoles, pro-orbicules
inside small plasma membrane invaginations and young orbicules
outside the tapetal cell surface (Plate V, 1).
The form of the microspores inside the callose envelope at this
stage is very characteristic and reminiscent of a ball covered with
blunt spines (Plate VI, 1). The optimal fixation of the material permits
observation of procolumella substructure; plenty of coils or circles
with a dark core, arranged in radial rows (for this see the central
procolumellae in Plate VI, 2 and a fragment of procolumella on the left
of it). Tiny branches of the free ends of the radial units, forming the
procolumellae, are especially distinctive above the left procolumella
(Plate VI, 2) by virtue of the transparent zones above the procolu-
mellar bundles. Detailed examination of the structure of the first
glycocalyx generation (marked by a star in Plate VI, 2) also show radial
arrangement of units of the fine network. Moreover, Plate VI, 4 clearly
shows that radial units of the second, high contrasted glycocalyx
generation are continuous with units of low-contrasted first glycoca-
lyx generation. The important feature of procolumella units is that
these units evidently extend through the plasma membrane into the
microspore cytoplasm (Plate VI, 3). This is especially evident in the
area of the cytoplasm above the oval vacuole. Another significant
feature is that the plasma membrane shows strong contrast imme-
diately at the base of procolumellae (Plate VI,2–5). Detail of the
substructure of the procolumella is clear in Plate VI, 4. The section is
slightly oblique and distinctly shows the separate units of this cluster
and even allows us to calculate an approximate number of units
alongside the diameter of procolumella: 5 or 6. It is clear from this that
the total number of tubular units in one columella is roughly from 20 up
to 40. The diameter of each is 50–70 nm and, what is significant, the
units have a light core and a dark “wrapper”(Plate VI,4).
Gradually the profile of the plasma membrane becomes rather
even. Procolumellae have grown and almost reached the callose
border. They acquire differential contrast, with highlights on the outer
surface (Plate VI, 5).
3.2.5. Late tetrad stage
Plate VII,1–2 show that the situation in the periplasmic space has
changed dramatically from the previous stage. Distinct young
columellae are observed on the surface of microspores in contrast to
the previous stage (Plate VI, 1). This critical change is evidently
associated with the beginning of sporopollenin (SP) accumulation. A
weak tendency for differential contrast of the procolumellae (which
has been observed at the previous stage, see Plate VI, 5) at this stage
increases by virtue of initial SP accumulations and “hollow”
columellae appear because of sharp difference in contrast between
outer “wall”of the columellae and their core part (Plate VII, 3). The
form of the columellae varies from cylindrical to X-like. It should be
stressed that the term “hollow”is used only as a description —to
emphasize uneven electron contrast of columellae. Actually, the core
part of every cylindrical or X-like columella is filled up with the same
radial units as their outer part. The detailed structure of a columella is
seen in detail in Plate VII, 4 (arrow), where radial units demonstrate a
spiral or circled substructure, the coils having light cores and dark
Plate I. Pretetrad and very young microspore tetrad stages. (see on page 215)
1. Sporogenous (SC) and tapetal (Ta) cells in the anther loculus. Sporogenous cells are angular. The cytoplasm of tapetal cells is more dense than that of sporogenous ones.
Cytomictic channels (arrows) connect tapetal cells. Scale bar =2 μm.
2. Cytomictic channel (arrow) between adjacent tapetal cells. Scale bar =1 μm.
3. The tapetal cell with an outgrowth intrudes between two microspore mother cells (MMC). The latter are enveloped with a thin callose layer. Scale bar=2 μm.
4. Metaphase in microspore mother cell. Arrow shows an invading tapetal cell. Scale bar =2 μm.
5. Very young tetrad stage. Microspores are covered with a thick callose envelope. Large nuclei occupy the central part of the microspores. Scale bar =2 μm.
6. A fragment of a very young tetrad microspore. The plasma membrane is smooth and lacks any signs of a cell surface coating. Scale bar =1 μm.
Ca=callose, Cy =cytoplasm, N =nucleus, PM =plasma membrane.
Plate II. Very young microspore tetrad stage in progress. (see on page 216)
1. Osmiophilic globules (arrows) in the periplasmic space between the plasma membrane of the microspore and callose envelope of a tetrad (a survey).
2. Osmiophilic concretions (arrows) on the surface of a tapetal cell (a survey).
3. Highly magnified dark globules in the periplasmic space (arrows), some of them have a jagged surface.
4. Endocytotic vacuole (asterisk) with engulfed portion of the microspore surface coating. Osmiophilic globules on the plasma membrane (arrows).
5. Osmiophilic globules (arrows) on the surface of invading tapetal cells (Ta).
6. A set of structures in the periplasmic space: globules of different size, spiral-like or tubule-like tiny structures (arrowheads), attached at the base to the plasma
membrane, globules in association with tiny tubules or spirals (arrows); unusual branching structures (a pair of arrows). All these structures are evidently different
mesophases of micelles and their transitory forms.
Ca=callose, Cy =cytoplasm, P =periplasmic space. Scale bars =1 μm.
Plate III. Young microspore tetrad stage. The beginning of the glycocalyx development. (see on page 217)
1. A survey of a part of the tetrad. The cell surface coating —the glycocalyx —is very thin. Scale bar =1 μm.
2, 3. The glycocalyx layer consists of roundish units with a dark core and light halo (arrows), with addition of radially oriented rod-like units. Coated vesicle in the
cytoplasm (arrowhead in 3). Scale bars =0.5 μm.
4. The plasma membrane is intensivelyinvaginated. Two generations of the glycocalyx are observed: fine network (star) and osmiophilic clusters of roundish units with
a dark contrasted core and less contrasted halo (arrows). Some of roundish units look branched (the left arrow), others are arranged into radial piles (the middle
arrow). Membrane-like structure in the periplasmic space (arrowhead). Plenty of ribosomes in the cytoplasm and dilated cysternae of endoplasmic reticulum (ER)
with osmiophilic contents. Scale bar =0.5 μm.
5. Clusters of roundish units on the surface of the plasma membrane (arrows). Network-like glycocalyx (star). The cytoplasm is full of dilated cysternae (ER), active
dictyosomes (D) and their vesicles (GV), ribosomes and small vacuoles. Scale bar= 1 μm.
6. Magnified fra gment of 5 showing clusters of roundish units with a dark con trasted core and lig ht halo, some of whi ch are arranged into radial piles. Note that
these units are interpreted as being spherical mice lles which are on the point of switching over to cylin drical micelles (radially oriented piles). Scale
bar=0.1 μm.
Ca= callose, CV=coated vesicle, N =nucleus, PS =periplasmic space.
214 N. Gabarayeva et al. / Review of Palaeobotany and Palynology 156 (2009) 211–232
Plate I (see caption on page 214).
215N. Gabarayeva et al. / Review of Palaeobotany and Palynology 156 (2009) 211–232
Plate II (see caption on page 214).
216 N. Gabarayeva et al. / Review of Palaeobotany and Palynology 156 (2009) 211–232
Plate III (see caption on page 214).
217N. Gabarayeva et al. / Review of Palaeobotany and Palynology 156 (2009) 211–232
218 N. Gabarayeva et al. / Review of Palaeobotany and Palynology 156 (2009) 211–232
binder elements. Cross sections through the columellae show rings
(that means that they are really cylindrical or close to cylindrical
structures with light-contrasted inner part) not pairs of columellae,
situated close to each other (Plate VII, 4, arrowheads; Plate VIII).
Central longitudinal sections through columellae shows that at this
ontogenetic moment their radial extent is equal to that of the
periplasmic space and is about 330 nm (Plate VIII, 1). The profile of the
plasma membrane has become even with rare invaginations. One such
concave area is seen in tangential section through the periplasmic
space where columellae have been cut at different planes of sec-
tion (Plate VIII, 2) and demonstrates different forms: cylindrical-like,
X-like, rings and circles (the latter correspond to sections of X-like
columellae in the region of “neck”) Images in Plate VIII, 3 and 4
correspond mainly to cross-sectioned columellae.
In the periplasmic space both generations of the glycocalyx persist
(Plates VII and VIII). At this stage the initiation of the tectum takes
place (Plate VII), and the ground for the future foot layer appears as
additional glycocalyx units on the surface of the plasma membrane.
These new components have a rounded form with a dark core and
light halo and are distributed, one by one or in clusters (Plate VIII,1, 2
and 4). The cytoplasm contains ER cysternae, active dictyosomes and
their vesicles, rare coated vesicles, vacuoles, and under-differentiated
mitochondria. Lipid-like droplets are often found in association with
the base of columellae.
4. Discussion
4.1. Introduction to gametophyte phase
The stage of sporogenous cells is preparatory for a very important
event in plant life —the reduction division, or meiosis. In most species
of seed plants meiosis is a synchronous process, and this synchroniza-
tion is coordinated by means of cytomictic channels between
sporogenous cells. A sporopolleninous sheet usually exists between
the endothecial cells and each tapetal cell, surrounding them from the
outer (distal) side. Rowley has called these individual sporopollenin
partial enclosures of tapetal cells “tapetal markers”(1982, 1988, 1999,
2000), because they really mark the initial position and size of every
tapetal cell. The tapetal markers can be of use in tracing the initial
position of tapetal cells and their initial form and size. These initial
parameters of tapetal cells can change through microspore ontogeny,
as was shown for Nymphaea colorata (Rowley et al., 1992a,b).
Outgrowths of some tapetal cells into the anther loculus also occur
in Trevesia at the microspore mother cell stage, although others retain
their palisade position.
4.2. Prelude to exine initiation
One of the main difficulties in ontogenetic studies (if there is not
an in vitro study, undertaken with pollen or stamen culture) is the
determination of the sequence of stages. The formation of structure is
a spatio-temporal process (Sattler, 1992), and an erroneous inter-
pretation of the sequence would mislead the interpretation of data
and observations. Many repeated fixations (for TEM) are needed to
ensure that the sequence of stages corresponds to reality. Discrete
stages in continuous development are surely imaginary things, but
they are useful to order our knowledge. Especially difficult is the
recognition of stages in the tetrad period, where no stable criteria exist
for stage definition.
As shown in the Results above, several transitory, preparatory
stages of exine establishment follow each other after meiosis and the
appearance of microspore tetrads with their thick callose envelope.
The main location of these events is the periplasmic space between
the plasma membrane and the callose envelope.
4.2.1. Is callose essential for exine development?
This crucial question has been asked many times by many authors,
and always with ambiguous conclusions. The consensus (for instance,
Dunbar, 1973; Barnes and Blackmore, 1986; Blackmore and Barnes,
1990; Hesse, 1995; Blackmore et al., 20 07), with the addition of recent
data on callose-deficient mutants (see below), suggests that callose is
necessary for normal exine development. This optimistic picture is
spoiled by the fact that most spore plants manage very well without
callose around their tetrads and tetraspores (for a review see
Gabarayeva and Hemsley, 2006), and heterosporous leptosporangiate
ferns have gone so far as to allow themselves to manage without the
tetrad period entirely (Lugardon, 1990). In fairness, it should be
remembered that most pollen walls are much more complicated in
their microarchitecture than most spore walls (exceptions, as usual,
exist) and, as we concluded earlier (Gabarayeva and Hemsley, 2006),
sporopollenin accumulation in spore walls is carried out sponta-
neously, without special receptor sites in the spore surface coating
(glycocalyx). Therefore, both features, the absence of callose and the
absence of specifically distributed sporopollenin receptors, are
coupled, hence a priori a relative simplicity of sporoderm structure in
spore plants.
In the anther locules of several cytoplasmic male-sterile petunia
lines, callose wall dissolution occurs earlier than normal as a result of
the premature appearance of callase activity (Izhar and Frankel,
1971 —cited by Worrall et al., 1992). It has been suggested that the
mistiming of callose wall degradation may be a primary cause of
male sterility in these lines. Worrall et al. (1992) have mimicked this
aspect in transgenic tobacco by engineering the secretion of a
modified β-1,3-glucanase from the tapetum prior to the appearance
of callase activity in the locule. The results obtained demonstrated
that premature dissolution of the microsporocyte callose wall was
sufficient to cause male sterility in transgenic tobacco. The authors
conclude (Worrall et al., 1992) that the aberrant microspore cell wall
of transgenic plants (noncompressed laminations, with globular
sporopollenin-like deposits above them) suggests the callose wall
is required for the correct formation and surface patterning of the
microspore exine.
In Arabidopsis at least 12 genes have been identified that are
involved in encoding callose synthase (CalS). It was demonstrated
(Dong et al., 2005) that one of these genes, CalS5, encoded a callose
synthase which is responsible for the synthesis of callose of meiocytes,
tetrads and microspores of tetrads. Knockout mutations of this gene
resulted in a severe reduction in fertility. Callose deposition in this
mutant was almost completely lacking. As a result, the exine was not
formed properly, affecting the bacculae and tectum structure. These
data suggest that callose has a vital significance in the establishment a
properly sculptured exine.
Plate IV. The completion of the young tetrad microspore stage.
1. Microspore of a tetrad (a survey). Its plasma membrane is periodically invaginated (arrows).
2. Magnified part of the microspore shown in 1. Arrows point to invaginative sites.
3, 4. Details of a microspore surface. The plasma membrane is essentially and periodically invaginated. Inside these vast invaginated areas, there are two generations of
the glycocalyx in the periplasmic space: a fine network (stars) and more contrasted, extended rod-like units, oriented radially and attached to the plasma membrane
(asterisks). Rod-like units, or tufts, are probably cylindrical micelles, self-arranged from spherical micelles. The same rod-like units are seen on the tops of
small evaginated sites, hence previously invaginated areas have undergone elevation at these points. The surface spherical micelles of the previous generation
(shown in III, 5) are probably partly engulfed in the process of endocytosis and seen in clusters in the cytoplasm (SM in IV, 4).
Ca= callose, Cy= cytoplasm, D= dictyosome, or Golgi stack, N= nucleus, RER=rough endoplasmic reticulum, SM= spherical micelles, V=vacuole. Scale bars= 1 μm.
219N. Gabarayeva et al. / Review of Palaeobotany and Palynology 156 (2009) 211–232
220 N. Gabarayeva et al. / Review of Palaeobotany and Palynology 156 (2009) 211–232
A striking case has been investigated by Anger and Weber (2006).
These authors demonstrated the absence of callose in the pollen
mother cells and tetrads of Arum alpinum resulting in no primexine
matrix (glycocalyx) and no ectexine. The thick endexine forms, but is
followed only by the appearance of polysaccharidic spines which do
not resist acetolysis. It should be stressed that a non-callosic space
between microspores in the tetrads exists, but it does not duplicate
the role of callose. Hence, a narrow gap between plasma membrane
and any enveloping “jacket”is not sufficient for normal exine
development, demonstrating that some special character of callose
is crucial. Another exciting point about microspore pattern formation
in Arum alpinum is that the outlines of spines are literally determined
by invaginations of an amoeboid tapetal plasma membrane (Anger
and Weber, 2006). These invaginations serve as a mould (template), a
principle considered to be rather widespread during pollen develop-
ment (Heslop-Harrison, 1968; Dickinson, 1976a, 1982; Gabarayeva,
2000). In essence, similar to the “principle of molding”is a “principle
of alternativity”(the inverse of a template) in which some structures
prevent the deposition of others (Gabarayeva, 2000; Gabarayeva and
Grigorjeva, 2003).
What is so special about callose that makes it essential? Callose
forms a semi-solid, elastic “jacket”around microspores and sets up a
constraint in the periplasmic space; this constraint is most probably
important for the establishment of the exine framework, glycocalyx,
or primexine matrix (see below). But why exactly callose? Why not,
in particular, a cellulose envelope which already covers the tetrad
as a whole from the time of microspore mother cell stage? The callose
(β-1,3-glucane), existing in a water-based medium of thecal fluid,
swells and represents a kind of homopolymer colloid (Hamm et al.,
2004). Being a swelling colloid (a gel—see Ball, 1994), the callose
envelope may behave actively in respect to the adjacent macro-
molecular layer (glycocalyx), promoting a structure-forming process
at the interface. In this respect, experimental work on simulation of
emulsification of an incompatible polymer interface (which has been
undertaken with technical aims) is very important (Hamm et al.,
2004). The authors have prepared a trilayer system of three different
polymers: amorphous homopolymer layer, liquid crystal (nematic)
homopolymer layer, and an amorphous-nematic copolymer at the
interface between them. The latter is usually used (in materials
technology) to increase the compatibility between the two former
layers. As a result, the homopolymers first induce a lamellar (“neat”
micellar) structure in the interfacial copolymer which, proving
unstable, changes towards interwoven bicontinuous micelles which
then fuse with both homopolymer blocks providing a mesoscopic
anchoring of the interface (the pattern of which resembles columel-
late-tectate exine patterning). This simulation provides an idea of
what is observed in the periplasmic space: a combination of callose
(amorphous homopolymer), glycocalyx (nematic or smectic crystals)
and emergent primexine pattern occurring as interfacial micellar
mesophases of three components (namely incompatible polymers
(amorphous callose and nematic liquid-crystalled glycocalyx) and
microemulsions of sporopollenin precursors).
The initial notion of the callose envelope as an impermeable
barrier was questioned long ago (Rowley and Dunbar, 1970;
Mascarenhas, 1975) and more recently (Gabarayeva, 1992; Scott,
1994), as it was shown experimentally that a number of substances
pass through callose (lanthanum nitrate —Gabarayeva and Rowley,
1994), and even large ions such as colloidal iron (Rowley and Dunbar,
1970; Rowley et al., 2003) and high molecular weight substances like
cerium ions/cerium perhydroxide precipitate (Rodríguez-García and
Majewska-Sawka, 1992) can pass through this layer.
4.2.2. Periplasmic space: arena for dynamic colloidal events
The gap between the plasma membrane of a microspore and the
internal surface of the callose cavity containing the microspore, the
periplasmic space is, essentially, where all major events connected
with exine development occur. In spite of the fact that the bulk of SP
accumulates in the post tetrad period, and endexine (as a rule) and
intine also appear after release of microspores from the tetrads, the
main design of exine pattern (which makes pollen grains of each
species what they are) occurs in the tetrad period in this narrow
periplasmic space. Strictly speaking, immediately after the completion
of meiosis and the formation of tetrads, there is no gap between
callose and microspore surface. Then a gap appears, and its width
increases gradually through the tetrad period, reaching several
hundred nm (up to 700–800 nm in Cabomba aquatica —Gabarayeva
et al., 2003) by the end of the tetrad period. It should be stressed that
the width of the periplasmic space is not constant alongside the
surface of a microspore as the profile of the plasma membrane is
always wavy, independent of any microarchitecture of the final exine
pattern. However, in the case of a future reticulate pattern, invagina-
tions of the plasma membrane are especially deep and regular.
The periplasmic space is filled with a water-based medium. This
is a natural consequence of the permeability of the tetrad callose
envelope which separates the microspores from the water-based
medium of locular fluid, containing polysaccharides and proteins,
during the tetrad period (Clément et al., 1998). Meanwhile, Golgi
vesicles deliver the first glycocalyx components (glycoproteins —
Rowley, 1971; Pettitt and Jermy, 1974; Rowley and Dahl, 1977; Pettitt,
1979) into the periplasmic space. Lipopolysaccharides are also
detected in the periplasmic space as constituents of the glycocalyx
(Rowley, 1975) and, up to middle tetrad stage, lipoid SP precursors
also appear in the same volume. All this makes the periplasmic space
an arena for colloidal relationships. The suggestion that the glycocalyx
has a colloidal nature and is established by self-assembly has been
offered (Gabarayeva, 1990, 1993, 2000). In reality, most solutions in
living creatures are colloidal, pure solutions being rather rare things.
The cytoplasm itself is a colloidal solution, which can turn from gel to
solid state and back (remember the pseudopodia of Amoeba).
First accumulations of Trevesia burckii, visible by TEM inside the
periplasmic space and on the plasma membrane, are dark contrasted
globules. Similar globules have been observed at this initial tetrad
stage in many species (Cosmos bipinnatus —Dickinson,1976a,b; Hesse,
1985;Triticum aestivum —El-Ghazaly and Jensen, 1987;Poinciana —
Skvarla and Rowley, 1987;Eucommia ulmoides —Rowley et al., 1992a,b;
Michelia fuscata, Magnolia delavayi —Gabarayeva, 1991;Stangeria
eriopus —Gabarayeva and Grigorjeva, 2002, Caesalpinia japonica —
Takahashi,1993;Nymphaeae colorata —Gabarayeva and Rowley, 1994;
Anaxagorea brevipes —Gabarayeva, 1995;Nymphaea mexicana —
Gabarayeva and El-Ghazaly, 1997;Illicium floridanum —Gabarayeva
Plate V. Middle tetrad stage.
1. A survey: tapetals cells (Ta) and a microspore (M) in an adjacent tetrad. The surface of the microspore is deeply invaginated. Proorbicules inside plasma membrane
invaginations (small arrows) and orbicules alongside the border of the tapetal cells (large arrows). Scale bar =1 μm.
2. Procolumellae on the top of the evaginated sites (arrows). They are clusters of radially oriented extended units (tufts) of 70 nm in diameter which are most probable
cylindrical micelles. Two generations of the glycocalyx in the periplasmic space: the first one (star) and the second (asterisks), the latter has lost the contact with the
plasma membrane. Scale bar= 0.5 μm.
3. Magnified part of 2. Substructure of procolumellae units–tufts is evident: the units have dark-contrasted cores and weak-contrasted halos, these correspond to
lipophilic interior and hydrophilic surface of cylindrical micelles. Electron transparent zones above procolumellae allow observation of the free ends of units
(arrowheads). First (star) and second (asterisk) glycocalyx generations, the latter appear as contrasted ribbons parallel to the plasma membrane. Scale bar =0.1 μm.
Ca =callose, N = nucleus, P= plastid, PM= plasma membrane, PS = periplasmic space, RER = rough endoplasmic reticulum, V = vacuole.
221N. Gabarayeva et al. / Review of Palaeobotany and Palynology 156 (2009) 211–232
222 N. Gabarayeva et al. / Review of Palaeobotany and Palynology 156 (2009) 211–232
and Grigorjeva, 2003; see also a review in Gabarayeva, 2000). In
Trevesia, as well as in Stangeria, tiny “cogs”are discernible on the
surface of these globules. The fact that such osmiophilic globules are
observed also on the surface of tapetal cells (Plate II, 5), and that the
first newcomers in the microspore periplasmic space (glycoproteins)
are diphilic, surface active substances (surfactants), the dark globules
on the microspore plasma membrane are most probably spherical
micelles. These are the first mesophase, self-assembling in a surfactant
solution of increasing concentration (Fig. 1 in Introduction), of a
captured lipophilic substance of tapetal/microspore origin in the
process of so-called solubilization (Mittal and Mukerjee, 1977), their
“jagged”surface being the hydrophilic heads of the glycoprotein
macromolecules. Significant images (Plate II, 6) assist in the under-
standing of what happens next. A mixture of dark globules, tiny
tubules or rods 40 nm in diameter, numerous chimerical dark globules
in connection with tiny tubules-rods resembling tadpoles, and even
branched tree-like structures all appear. We think that, at this
ontogenetic moment, a kind of diffuse glycocalyx occurs, in which
the distribution of the material is uneven through the width of the
periplasmic space. As a result, “hybrid”micelles appear; spherical
at the plasma membrane, with rudiments of cylindrical micelles
protruding from the former; tree-like images may show semi-
developed cylindrical micelles (tufts, by Rowley and Flynn, 1968;
Rowley, 1990) with their loops of binder spirals, cut transversally. It is
perhaps meaningful that 15 years ago, studying exine development in
Nymphaea colorata, we observed numerous tadpole-like and tree-like
structures in the periplasmic space at the early and middle tetrad
stage and supplied micrographs with sketches to facilitate their
interpretation. We also observed, in periplasmic space, short lamellae
with a central white line (Gabarayeva and Rowley, 1994). Now
we suggest that all these types, seen in periplasm of Nymphaea and
Trevesia, are different mesophases of a micelle system, existing in the
periplasmic space: spherical, cylindrical, lamellar (so-called “neat”)
micelles with a typical gap between laminae, and cylindrical micelles
in the process of self-arrangement (tadpole-like images).
The interpretation is that a lipoidal osmiophilic substance from the
tapetum passes through the callose to the periplasmic space of the
microspores (or is derived from the microspore cytoplasm) and
undergoes solubilization within the spherical and cylindrical micelles
of the glycocalyx; then, in the course of endocytosis (which alternates
or is concurrent with a exocytotic wave of glycoprotein excretion), it
further passes into the microspore cytoplasm for re-cycling.
The following structures, which come forward to the periplasmic
arena (Plates III and IV), are all the same familiar members of transi-
tional micellar mesophases. These include roundish entities with a
dark core and light contrasted halo (spherical micelles, Plate III, 2, 3),
contrasted lines (membranes in self-assembling process, Plate III, 4),
and fine networks (material of the first generation of the glycocalyx,
which represents a kind of molecular and supramolecular “soup”of
spherical and cylindrical micelles, their remnants and their progeni-
tors, depending on the width of the periplasmic space in the loci. This
width is variable because of the dynamic-invaginative character of the
plasma membrane (plasma membrane micromovements —Gabar-
ayeva, 2000; glycocalyx generations —Gabarayeva and Grigorjeva,
2002). The colloidal mixture constantly suffers variable constraint,
which is the result of the variable volume of the periplasmic space. An
interplay of contractile microfilaments (MFs), attached to the plasma
membrane, might be the cause of this variability. Finally, dark
contrasted units on the plasma membrane are clusters of spherical
micelles (Plate III, 4, 5), the latter gradually self-arrange into
cylindrical ones, attached to the plasma membrane (Plate IV, 3, 4).
Clusters of roundish particles in the cytoplasm (apparently spherical
micelles, seen in Plate IV, 4) may consist of those spherical units,
which have been engulfed in the process of endocytosis.
These pre-exine stages have been transitive, preparatory ones
which then pass initiative to the next ontogenetic events of exine
establishment.
4.2.3. Plasma membrane, its glycocalyx and tensegrity
4.2.3.1. Plasma membrane. A transitional stage with shallow plasma
membrane invaginations (shown in Plate IV) is accentuated in the
next stage which actually shows reticulate pattern of the primexine
apparently before SP accumulation (Plate V). Indeed, the plasma
membrane is deeply and regularly invaginated. The glycocalyx on the
tops of undulated plasma membrane is prominent as bunches of radial
cylindrical micelle-tufts. This differs from the portions of the
glycocalyx inside the invaginations, where the glycocalyx is detached
from the plasma membrane and never develops further. Regularly
convoluted plasma membrane is a general developmental feature of
all species with a reticulate exine pattern. With the attainment of the
middle tetrad stage, all of the main events in the establishment of a
future exine pattern have occurred. What, however, is the mechanism
which forced the plasma membrane to take such a spatial form? Why
are some portions of the glycocalyx (on the top sites) promoted in
growth, whereas others are stopped? Before suggesting the answer, it
would be reasonable to review, in short, what is known about the
plasma membrane, its extracellular matrix (glycocalyx), and the
cytoskeleton–cell-wall continuum.
Membranes represent the main structural component for the
complex architecture of biological systems. Membranes composed of
amphiphilic molecules are highly flexible surfaces that determine the
architecture of biological systems and provide basic structural
elements for complex fluids such as bicontinuous structures, including
unusual sponge-like phases. Physical theories describe the conforma-
tional behaviour of membranes such as preferred shapes, shape
transformations and shape fluctuations, adhesion and unbinding of
membranes (Lipowsky, 1991).
Lipid bilayers often form layered structures: multilamellar vesicles
(spheres), cylindricals or oriented stacks. Other possible structures
are 3-D periodic surfaces which divide a space into two interwoven
labyrinths or subvolumes, giving rise to a bicontinuous phase. Such
semi-crystalline states of bilayers are known for amphiphilic
molecules and, if with minimal interface and equal subvolumes, are
called periodical minimal surfaces of Schwarz (Scriven, 1977). In plant
cells such structures are often seen in the cytoplasm as prolamellar
bodies (Gunning and Steer, 1986), as chain-mail aggregates of ER
(Gabarayeva,1986, 1987) or as prolamellar bodies in plastids (Caredda
et al., 1999). There is little doubt that granae of thylakoids in plastids
are organised on the basis of lamellar micelles. The association of
Plate VI. Middle tetrad stage.
1. Microspore in the form of a ball with blunt spines (survey). Scale bar = 1 μm.
2. Substructure of procolumellae seen as coils of radially oriented tufts, united into bundles on the surface of plasma membrane evaginations, is evident (arrows). The free
ends of these tuft-micelles are distinctive on the tops of procolumellae on the electron transparent background. Scale bar =0.5 μm.
3. Radially oriented tuft-units of procolumellae (arrows) go through the plasma membrane into the cytoplasm (arrowhead). Scale bar =0.5 μm.
4. The initial generation of the glycocalyx (star) shows radial arrangement of units of the fine network. Moreover, radial units of the second, high contrasted glycocalyx
generation (asterisk) are continuous with initial units. The plasma membrane is strongly contrasted exactly at the base of procolumellae, shown also well in 2, 3 and 5.
Scale bar=0.5 μm.
5. The profile of the plasma membrane is flatter than in earlier stage. Procolumellae have grown and almost reached the callose border (arrows). They acquire differential
contrast: highlights are seen on the outer surface. Scale bar =0.5 μm.
Ca=callose, Cy =cytoplasm, PS =periplasmic space. Stars =first glycocalyx generation, asterisks =second glycocalyx generation.
223N. Gabarayeva et al. / Review of Palaeobotany and Palynology 156 (20 09) 211–232
224 N. Gabarayeva et al. / Review of Palaeobotany and Palynology 156 (2009) 211–232
spherical and lamellar (“neat”) micelles have been most probably
described in plastids as so-called membrane-particle associations
(Dickinson and Willson, 1983), where the lipid fraction and
carbohydrates meet, and were observed in plastids of tetrad
microspores of Anaxagorea (Gabarayeva, 1995).
Surfactant molecules in mixtures of immiscible fluids often
assemble into membranes. Biomembranes are typically fluid, which
means that the molecules can diffuse rapidly within the membrane.
Fluid membranes could be highly convoluted or crumpled. Biomem-
branes often contain two-dimensional protein networks, such as a
spectrin network on the plasma membrane of red blood cells; these
networks can be regarded as “fishnets”of fixed connectivity. The mesh
size of protein networks is about 100 nm (Elgsaeter et al., 1986).
There can be two states for several fluid membranes in a system:
exhibiting non-binding and adhesive. The latter takes place when two
or more membranes stick together. The transition between these two
states is temperature-dependent. Non-binding and adhesion transi-
tions of fluid membranes have been found in experiments with sugar-
lipid membranes, where these membranes were unbound and exhibit
string undulations, appearing in light microscope as a thick fuzzy line
(but as a bundle of separated membranes in TEM) at a definite
temperature. If the temperature decreases, the membranes suddenly
form an adhesive state, sticking together along vast regions and
leaving some fragments in an unbound state, where membranes
disperse as a fan (Lipowsky, 1991). All this fits very well with experi-
mental results of Rowley et al. (2003) on transfer of material through
exine: large transitory non-apertural bulges of the lamellar endexine
of living Epilobium microspores close within 10 min, if the microspores
have been put in cold (4 °C) phosphate buffer and observed in light
microscope. Rowley concluded long ago (1976) that the exine was
sufficiently plastic during pollen development for great alterations in
shape. Now we understand this phenomenon on a physico-chemical
level such that the fluidity of biological membranes makes them
plastic and subject to adhesion or separation, dependent upon
different factors, including temperature.
The point of interest focusses, as earlier (Heslop-Harrison, 1972;
Gerasimova-Navashina, 1973), on the mechanism of exine pattern
determination (Gabarayeva, 1990; van Uffelen, 1991; Hemsley et al.,
1992; Collinson et al., 1993; Gabarayeva, 1993; Hemsley et al., 1994;
Southworth and Jernstedt, 1995; Hemsley et al., 1996, Hemsley et al.,
1998, 2000; Borsh and Wilde, 2000; Gabarayeva, 2000; Griffiths and
Hemsley, 2001; Hemsley et al., 2004; Gabarayeva and Hemsley, 2006;
Hemsley and Gabarayeva, 2007). In species with a reticulate exine the
initial undulating conformation of the plasma membrane plays,
without doubt, an important role in the pattern establishment (see,
for example, Takashashi, 1986, 1991 —although the author did not
mention in which way). Such movements of the plasma membrane
(macromovements —Gabarayeva, 2000), once having occurred, then
stop, and the plasma membrane retains its wavy profile for some time,
usually through the tetrad period. But what is the reason for the
appearance of the plasma membrane invaginations? Are these depres-
sions formed as the result of passive pushing inside of the areas of the
plasmalemma in sites where the glycocalyx grows and increases in
volume? If so, these sites should be predetermined by the plasma
membrane itself. Or do these invaginations appear as a consequence of
pulling inside by contractile microfilaments (MFs), anchored to the
nucleus envelope? This question is perhaps reminiscent of the well-
known unresolved question about the primacy of hens or eggs. It
could be a case of coupled mechanisms, or coupled modes of one
mechanism, providing the intrinsically robust morphogenesis (Good-
win et al., 1993). The spatial and temporal correlations were described
between the distributional pattern of cortical microtubules (MTs) and
coarsely reticulate exine patterns in Vigna vexillata (Muñoz et al.,
1995). These results did not concur with earlier studies on Lilium
(Dickinson and Sheldon, 1984; Sheldon and Dickinson, 1986), where
the authors concluded that MTs did not correlate with exine pattern in
developing microspores. Working on microspore development of
Vigna unguiculata,Southworth and Jernstedt (1995) concluded that
microtubules did not determine exine pattern, on the contrary, the
developing pattern rearranged microtubules. The authors believed
that the pattern formed was a response to tencile and rigid properties
of the cytoskeleton (tensegrity —see below) and to osmotic pressure
in the microspore, balanced against the pressure and volume of
newly secreted matrix. This model for pollen wall patterning includes
a self-patterning idea and suggests that semi-stiff MTs bend or move
to areas of plasma membrane protrusions, allowing greater indenta-
tion of the plasma membrane in other regions (Southworth and
Jernstedt, 1995). But then the question arises: what causes the initial
indentations?
Sheldon and Dickinson (1983) suggested purely physical phenom-
ena in establishment of exine patterning, with the participation of
coated vesicles, inserted randomly into the plasma membrane, where
the deposited protein, being hydrophilic, initially floats as circular
plates within the fluid lipid part of the plasma membrane. When
sufficiently accumulated, these plates become tightly-packed and self-
assemble into a hexagonal pattern (like bubbles within a foam) and
later the pattern would become stenciled on the plasma membrane
and baccules form along lines specified by the interfaces between the
plates. This hypothesis seems to us very tempting because of its
physico-chemical simplicity and its principle similarity to colloid
micelle behaviour, the more so because hexagonality is a constant
pattern feature (Scott, 1994), inherent in one or another stage of the
tetrad microspore and not restricted to reticulate exine patterns
(Gabarayeva, 1993, 200 0). One circumstance however, seems confus-
ing and that is the rarity of coated vesicle images observed in
microspore development.
Perhaps significantly, Robenek (1980), studying isolated protoplasts
of the callus in Skimmia japonica by the method of freeze-aching and
TEM, discovered that while protoplasts began to regenerate their
envelopes an identical pattern of particles were observed on both
plasma membrane and ER cysternae (which were in contact with
each other), and these patterns were mostly hexagonal. The author
(Robenek, 1980) supposed that these particles were protein or
lipoprotein molecules, and they appeared to transfer from the surface
of ER to the plasma membrane into the special hybrid lipid membrane
bilayer. Three points seems to be important from this work; the idea of
imprinting of a pattern, the hexagonal (reticulate) nature of the pattern,
and the expected chemical composition of particles forming the pattern.
Plate VII. Late tetrad stage. The beginning of sporopollenin accumulation.
1–2. Surveys of tetrads with well-pronounced columellae, some of which look “hollow”(arrows). Scale bars = 2 μm.
3. Longitudinal section through young “hollow”columellae showing their cylindrical or X-like form and a very uneven accumulation of sporopollenin (arrows). Large
spiral coils around one of the columellae (arrowhead) of which a detail is shown in the inset (white arrowheads). Scale bar =0.5 μm.
4. A section through differently bent columellae: longitudinally sectioned columella (arrow) and columellae which are transversely sectioned or nearly so
(arrowheads). The inner, core part of the columellae is filled with the same radial units as the outer part, but almost lacks contrast. The fine substructure of
columellae is prominent: radial units demonstrate spiral or circular substructure, the coils having lightcores and dark binder elements. These substructural units are
evidently cylindrical micelles. The image of the columella with large spiral coils on the surface (3, inset, white arrowheads) suggests a possible re-arrangement of a
bundle of cylindrical micelles into one huge supramicelle. Stars in 3 and 4 show first generation of the glycocalyx; white asterisks show the second generation. Scale
bar=0.25 μm.
Ca=callose, D =dictyosome, ER =endoplasmic reticulum, GV =Golgi vesicles, N =nucleus, PM =plasma membrane, PS =periplasmic space.
225N. Gabarayeva et al. / Review of Palaeobotany and Palynology 156 (20 09) 211–232
226 N. Gabarayeva et al. / Review of Palaeobotany and Palynology 156 (2009) 211–232
In earlier papers (Gabarayeva,1990, 1993, 2000) two pathways for
exine pattern determination have been suggested: (1) the introduc-
tion of a ready template for the future pattern into the periplasmic
space for the glycocalyx (primexine matrix) formation by imprinta-
tion of the sporophytic information from the plasma membrane (the
latter receives it during contact time with ER and/or other
cytoplasmic organelles, or bears the code for the pattern formation
from the time of premeiosis); (2) the disordered, random introduction
of enzymes —catalysts of SP polymerisation (SAPs) into self-
assembling 3-D network of the glycocalyx, with subsequent self-
insertion of SAPs, resulting from the balance of weak and strong
molecular SAP-glycocalyx and SAP–SAP interactions. As we see it now,
the first pathway suggests the conversion of 2-D information (coded
on the surface of the plasma membrane) into 3-D one. Such a problem
exceeds the boundaries of the purely biological field and appeals to
geometry, but could be suggested as unfolding of a “flat”pattern to a
“volumetric”, with the obligatory presence of a temporal parameter. A
suggestion is found in Blackmore et al. (2007) who has advanced an
opinion that it is the plasma membrane that should be regarded as
mediating a pattern. The latter will be generated if the process takes
place at different times in different places on the microspore surface
(italics are ours). The second suggestion, based mainly on physico-
chemical regularities, was developed and extended, accenting the
properties of SP emulsions (Hemsley et al.,1992; Collinson et al., 1993;
Hemsley et al., 1994), confirmed by experimental modelling of exine
self-assembly (Hemsley et al., 1996, 1998; Griffiths and Hemsley,
2001; Hemsley et al., 2004) and recently proposed as a model of exine
development as a micelle system (Gabarayeva and Hemsley, 2006;
Hemsley and Gabarayeva, 2007).
To our satisfaction, we have found confirmation of our ideas in the
recent paper of Nishikawa et al. (2005) who worked with alleles of the
Arabidopsis CALS5 gene, dissecting the role of callose synthase in
pollination. Unlike strong CALS5 alleles (cals5-1 and cals5-2) that
caused pollen degeneration and sterility (Dong et al., 2005),
Nishikawa et al. (2005) identified fertile mutants cals5-3, cals5-4
and cals5-5. A strong defect shows cals5-3: the tetrads completely lack
a peripheral callose layer and have little callose between microspores;
the peripheral callose layer is also missing from cals5-4 and cals5-5,
but they have callose between microspores. All three mutants have
aberrantly patterned exine. Regretfully, the authors give illustrations
only of mature pollen grains, so it is impossible to see developmental
stages of the pollen wall and judge the degree of primexine matrix
disturbance. Most interesting are the conclusions of these authors that
both the extracellular callose layer and the intracellular factors which
nucleate SP deposition are required for exine formation. Callose could
trap primexine subunits, increasing their local concentration, then
these subunits self-assemble into a scaffold, nucleated by intracellular
components for SP accumulation. The possibility that callose provides
a physical support for primexine assembly, is also suggested. These
conclusions are very close to what we proposed in “micelle language”
i.e. a gradual increasing of glycoprotein concentration in the peri-
plasmic space, leading to appearance of several successive meso-
phases of micelles, resulting in self-assembly of the glycocalyx
(primexine matrix), and subsequent distribution of SAPs —the sites
of SP nucleation (Gabarayeva and Hemsley, 2006; Hemsley and
Gabarayeva, 2007; data in this paper). As for the physical support,
exerted by callose for primexine assembly, it has aspects in common
with tensegrity ideas (see below). However, one of the authors
conclusion (Nishikawa et al., 2005) sounds more than strange; they
inferred that a structured exine is not required for pollen develop-
ment, viability or fertility (on the basis that the examined mutants
develop pollen tubes). It is possible that such a conclusion has arisen
from the fact that the authors studied pollen–stigma adhesion in self-
pollinated pistils: pollen without an exine will develop in self-
pollinating pistils, but lack of a wall is lethal as soon as pollen is
released beyond the flower. If a structured exine were not required for
pollen development, it would never be preserved in evolution.
4.2.3.2. Glycocalyx. The plant cell wall (especially, to our mind,
microspore glycocalyx, or primexine matrix) can be considered
homologous to an animal cell extra-cellular matrix (ECM) and in
this context it would be relevant to regard the idea of the
cytoskeleton–cell-wall continuum (Wyatt and Carpita, 1993)asa
real system, involved in the appearance of periodic plasma membrane
invaginations and the location of procolumellar bundles of the
glycocalyx on top of protrusions. (Wyatt and Carpita reasonably stress
that the term “extracellular matrix”should be replaced with the term
“exocellular matrix”, because the cell surface coating —glycocalyx —is
an integral part of the cell and not something outside the cell). The
ECM of animal cells is composed of glycoprotein, proteoglycans and
glycosaminoglycans (Adams and Watt, 1993), and the ECM (glycoca-
lyx) composition of plant cells, in particular of microspores, is similar:
mainly glycoprotein, with the addition of lipopolysaccharides and
lipids (Rowley, 1971; Pettitt and Jermy, 1974; Pettitt, 1979). The ECM
plays a role in regulating the differentiated phenotype of cells, and can
also act as a physical barrier or selective filter to soluble molecules.
The authors (Wyatt and Carpita, 1993) conceive the cytoskeleton and
ECM as an interactive scaffold for the perception and transduction of
positional information and suggest that a cytoskeleton–ECM con-
tinuum may be a common feature for all eukaryotic cells —plants,
fungi and animals. Current knowledge points to an interaction of the
cytoskeleton and ECM via trans-membrane proteins of the plasma
membrane. For example, the pattern of association of cortical
microtubules with the plasma membrane reflects the pattern of
cellulose deposition in both primary and secondary walls. This is
because inhibition of cellulose biosynthesis prevents normal function-
ing of actin filaments, etc. (see Wyatt and Carpita, 1993 and literature
included). We have observed MTs and MFs during exine development
in many species (despite not having used special methods to
distinguish them), for example, in Anaxagorea brevipes and Illicium
floridanum (Gabarayeva, 1995; Gabarayeva and Grigorjeva, 2003).
Radially oriented MF are distinctly seen running down from the
plasma membrane portion, underlining procolumella, to the cyto-
plasm in Poinciana microspores (Skvarla and Rowley, 1987, Fig. 16)
and in Plate I, 3 in Gabarayeva et al. (submitted to the same issue —the
second part of Trevesia exine development). In Trevesia exine
ontogeny, fragments of MTs are clearly seen in Fig. VI, 3 (arrowhead),
and procolumella (arrow) appear rooted in the cytoplasm via the
Plate VIII. Late tetrad stage. The beginning of sporopollenin accumulation (cross sections on the whole).
1. A central radial section through columellae. These have a cylindrical form. Newly formed glycocalyx units on the surface of the plasma membrane (white arrowheads).
Lipid droplet at the base of a columella (black asterisk). First generation of the glycocalyx (star); second generation (white asterisk). Scale bar=0.5 μm.
2. Tangentially sectioned concave area through the periplasmic space where columellae had been cut at different planes of section and demonstrate different images:
cylindrical-like (black arrow), X-like (white arrow), rings (double arrowhead) and circles (black arrowheads), the latter correspond to sections of X-like columellae in
the region of “neck”. First (stars) and second (white asterisks) generation of the glycocalyx. New glycocalyx units at the plasma membrane (white arrowhead). Scale
bar=0.3 μm.
3. Cross-sectioned columellae looking like rings (arrows). Scale bar =0.3 μm.
4. Young columellae, sectioned in different planes: longitudinally sectioned (arrow), bent (double arrowhead) and cross sectioned (arrowheads). Small arrows point to
new glycocalyx units on the plasma membrane. Stars and asterisks in 3 and 4 show first and second generations of the glycocalyx. Scale bar= 0.3 μm.
Ca =callose, CV = coated vesicle, Cy = cytoplasm, PM= plasma membrane, PS = periplasmic space.
227N. Gabarayeva et al. / Review of Palaeobotany and Palynology 156 (2009) 211–232
plasma membrane, a common observation in ontogenetic studies.
Careful examination allows us to distinguish traces of cytoskeletal
bundles through Trevesia microspore cytoplasm, but they are not
prominent.
Another important feature of the cytoskeleton–ECM continuum is
adhesion, a physical attachment of the plasma membrane to the wall
(in our case —to the glycocalyx). Local adhesion (or its absence) plays
an important role in functioning of cytoskeleton–EMC continuum. It
was shown that extracellular matrix receptor, integrin β
1
induces focal
adhesion formation and supports a force-dependent stiffening
response, whereas non-adhesion receptors did not (Wang et al.,
1993); tensegrity models mimic this response, suggesting that
integrins act as mechanoreceptors and transmit mechanical signals
to the cytoskeleton. As described in Results, from the beginning of the
middle tetrad stage (Plate V) the second generation of the glycocalyx
(marked elsewhere by asterisks) looses contact with the plasma
membrane and becomes detached from it in invaginated sites,
between procolumellae-bearing protrusions. From this point in
ontogeny no new portions of the glycocalyx appear in these
invaginative sites, and no sporopollenin (SP) accumulates through
these “rejected”glycocalyx areas. Whether it is the stopping of the
glycocalyx material delivery that is the reason for the conservation of
the previously-formed glycocalyx layer or, on the contrary, the lost of
the connection between the glycocalyx and the plasma membrane
prevents further building-up the glycocalyx, is not clear. In any case,
self-assembly of the glycocalyx cylindrical micelles ceases in areas
inside plasmalemma depressions, whereas at pinnacles it continues.
The images shown in Plate VI, 2 and 4 deserve special attention.
They clearly show that the first and second glycocalyx generations
(marked by stars and asterisks) represent a continuum of radially
oriented cylindrical micelles (tufts). Moreover, there is little difference
between procolumellar units and those between them, other than in
the degree of apparent contrast. Another important point is that radial
cylindrical micelles (tufts), gathered in ensembles on pinnacles of the
plasma membrane protrusions, initially have dark-contrasted cores
and light halos (see, for instance, Plates V, 3 and VI, 2). Somewhat
later, some of them have light cores and dark halos (Plate VI, 4) and
this feature becomes universal up to the end of the tetrad period. This
wide-spread phenomenon —the reversal of the contrast, described
earlier (Dunbar and Rowley, 1984; Rowley and Claugher, 1996;
Gabarayeva et al., 1998)—is most probably connected with inversion
of so-called direct micelles (hydrophobic core, hydrophilic surface-
binder —see scheme in Fig. 1, d) to reverse micelles (hydrophilic core,
hydrophobic surface-binder —see Fig. 1, d'). Such a “turning inside
out”occurs because of gradual change of chemical content of the
supporting medium in the periplasmic space as lipoid SP precursors
enter inside the periplasmic liquid.
4.2.3.3. Tensegrity. The tensegrity paradigm demands special con-
sideration. The term (tensional integrity) has come from architecture,
where it means a property of framed structures, based on the balance
of continuous tension and discontinuous compression in combination
such that every element of the structure works to maximum
effectiveness and economy. It was R. Buckminster Fuller (1965) who
first spoke of tensegrity. His student, a sculptor K. Snelson, for the first
time constructed a tangible model based upon this system of tensile
organization. Such architectural constructions are comprised of an
assemblage of compression-resistant struts (e.g., soda straws, bones,
wood dowels) that do not physically touch one another but are
interconnected by a continuous series of tension elements (e.g., elastic
threads, muscles, monofilament lines), so they are pulled up and open
rather than compressed in place (Ingber and Jamieson, 1985). These
tensegrity models if pressed to a substrate and flattened literally jump
up from the surface and restore their form when free. Amazingly,
these models predict precisely a wide variety of CSK patterns,
including “geodomes”. The correspondence between tensegrity
models and hand-drawn depictions of published light and TEM
micrographs, showing the triangulated arrangement of MFs within a
CSK geodome, is striking (Ingber, 1993). If such a model were
anchored at multiple points to a malleable substratum, it would
spontaneously retract, pull its attachments together and compress the
underlying foundation into folds (Ingber and Jamieson, 1985; Ingber,
1993). The authors (Sims et al., 1992; Ingber, 1993), working with a
cell culture, concluded that living cells acted in a nearly identical
manner. They flattened when attached to a highly adhesive substrate
but detached and became rounded when their ECM anchors were
enzymatically removed. They physically pulled elastic substrata into
“compression wrinkles”, and spontaneously contracted malleable ECM
gels (Ingber, 1993 and literature cited). We used italics in the previous
phrase to emphasise a striking similarity of changes described and
those we observed in developing Trevesia microspores around the
middle tetrad stage, i.e. a deeply, periodically invaginated microspore
surface with a contracted, wrinkled glycocalyx (ECM gel), non-adhesive
(detached) plasma membrane depressions and adhesive pinnacles
(Plate V, 2, 3). This partly “rejected”glycocalyx–ECM is distinctly seen
at low magnification (Plate V,1;Plate VI, 1), being separated from the
microspore plasma membrane by an electron transparent gap. The
nucleus, to some extent, mimics the microspore cell form (reminis-
cent of a hedgehog which has bristled-up). Such forms of tetrad
microspore, typical for all species with reticulate exines, might occur
as an interplay of osmotic pressure (in the locked space inside the
rather rigid callose jacket) and the tensegrity forces, innate to any cell.
Large-scale changes in cell and nuclear shape result from the action of
mechanical tension that is generated within the cytoskeleton via an
actomyosin filament sliding mechanism and transmitted across
integrin receptors of the plasma membrane. It is physically resisted
by immobilized adhesion sites within the extra(exo)cellular matrix
(Sims et al., 1992). These authors, working with an endothelial cell
culture, have found that rapid and coordinated changes of cell,
cytoskeletal, and nuclear form result when the cellular force balance is
altered. They used a variety of methods to shift forces from the ECM
onto internal cytoskeleton struts (by enzymatic removal of adhesions
between ECM and plasma membrane integrines), and by increasing
cytoskeletal tension above normal level (by adding ATP). The authors
observed similar (but different speed) changes in both cases (the
extended cell processes pulled back, following cytoplasmic and
nuclear retraction) and found that cytoskeletal tension must be
applied to integrins to produce cell and nuclear retraction. All of these
findings support the concept that integrins of the plasma membrane
and ECM normally control both cell and nuclear shape by physically
resisting cytoskeletal tension (Ingber, Jamieson, 1985; Sims et al.,
1992). It is highly possible that other cytoskeletal elements, such as
MTs and intermediate filaments (IFs), also play a role in maintenance
of force balance. Recent works provide strong evidence to support the
use of tensegrity by cells (see reviews by Ingber, 2003a,b). Tensegrity
includes two key determinants: architecture (3-D geodome construc-
tion) and pre-stress. The former triangulate their structural members
and orient them along geodesics (minimal paths); the latter hold
their joints in position as the result of “pre-existing tensile stress”
(pre-stress). The cellular tensegrity model proposes that the whole
cell is a pre-stressed tensegrity structure. In other words, cells
generate their own internal tension or pre-stress in the actin
cytoskeleton (MFs pull plasma membrane inside), which is balanced
by internal microtubule struts (MTs do not allow plasma membrane to
sag inside) and external ECM adhesions (these focal adhesions
represent discrete points of cytoskeletal insertion on the ECM
analogous to muscle-insertion sites on bones (Ingber, 2003a). It is
shown experimentally that when the ECM is carefully removed from a
cell, it partially retracts. Cells cannot preserve their spread form in the
absence of their ECM adhesions (which are clustered integrin
receptors and cytoskeleton-coupling proteins —see literature cited
in Ingber, 2003a). Such experiments have shown that to support cell
228 N. Gabarayeva et al. / Review of Palaeobotany and Palynology 156 (2009) 211–232
spreading, isolated regions of the extracellular matrix, located
between focal adhesions, must resist local compression produced by
the shortening of each internal stress fiber. It is for this reason that
cells can pull a flexible substrate up into “compression wrinkles”
between their localized adhesions (Harris et al., 1980). Cells must
contain some internal elements that resist inward-directed cytoske-
letal forces in order to extend outward and this is a key feature of
tensegrity architecture. Radially oriented MTs play this role in cells.
Moreover, for tissue morphogenesis, it has been shown (Ingber, 1993)
that thinning of the ECM scaffold will result in an increase in its
mechanical compliance and cause a local cell distortion through the
action of tractional forces exerted by surrounding cells. The con-
sequence is an invaginated region (Ingber, 2003b). The same
phenomenon would occur on the cell level. Thinning of the ECM
would result in an increase of its compliance and cause a local
depression because of the action of inward-directed forces imposed by
contractile MFs on the plasma membrane in a pre-stressed cell.
It is highly probable that analogous processes takes place in
microspore tetrad development. Once balanced at early tetrad stage,
when the “bare protoplast”of a microspore suffers pressure inside the
semi-rigid callose jacket, the cell takes a specific concave–convex form
(Plate I,5;Plate III, 1). At this initial stage microspores most probably
are not pre-stressed because they are restricted and supported from
all sides by the callose envelope. However, the cellular force balance
becomes altered and shifted as soon as the glycocalyx (ECM) appears
(Plate III,2–4): The ECM molecules probably produce cell shape
changes via binding to the cell surface by integrin receptors which
span the plasma membrane and interconnect with actin MFs inside
the cell. As a result of the shifted balance and interplay between the
oppositely directed forces of inward-directed MFs pulling and
outward-directed resistance of ECM–glycocalyx, the form of the cell
surface changes such that the microspore no longer occupies the
whole space inside the callose. Unexpectedly, the absolute volume of a
microspore in a tetrad decreases, as demonstrated in a stereology
study (Gabarayeva and Grigorjeva, 2002). This effect is clearly seen in
Trevesia in Plate V, 1. As soon as a microspore “feels free”of the callose
substrate, it retracts even more (Plate III, 4), possibly because free or
partly detached cells exhibit retraction (see above). By this ontoge-
netic point, or slightly later, new and different glycocalyx accumula-
tions (initial cylindrical micelles) appear locally on the plasma
membrane (shown by asterisks in Plate IV) which introduce new,
most powerful sites of resistance to microspore pre-stressed MFs and,
to some extent, overcome it. As a result, these sites become pinnacles
of the plasma membrane and focal adhesions of the ECM–glycocalyx.
Then, thinned regions of the extracellular matrix (glycocalyx, which
has ceased to grow) located between focal adhesions at the plasma
membrane pinnacles (Plate V), being more compliant, cannot resist
in full measure the inwardly-directed tractional forces of the pre-
stressed MFs, and thus the plasma membrane invaginations and ECM-
compression wrinkles. Therefore, the cell itself takes on the form of an
angled entity, the bristled hedgehog referred to above (Plate V, 1; VI,
1) because of unsupported portions of the plasma membrane (seen as
invaginations) with the detached glycocalyx (ECM) between pinna-
cles of the CSK, distributed underneath the plasma lemma protru-
sions, and adherent to the glycocalyx. Indirect confirmation of the
tensegrity-caused appearance of the intensive, regularly invaginated
microspore surface in the tetrad period (which prepatterns the future
reticulate exine) is the fact that only species with a reticulate pattern
have, in their ontogeny, a period (about middle-late tetrad)
characterized by a glycocalyx arrested in growth and therefore
thinned in invaginations and promoted in growth on pinnacles. Our
reconsideration of our pollen wall ontogeny studies on this special
feature confirms the fact (see, for example, data on Borago —
Gabarayeva et al., 1998;onIllicium and Schisandra —Gabarayeva,
Grigorjeva, 2003). Bissel et al. (1982) suggested that ECM, as a
multivalent ligand with a distinct periodicity, may cluster surface
growth factor receptors or trans-membrane molecules which in turn
alter intracellular filament nucleation sites and polymerization (the
concept of “extended cytoskeleton”). Thus, one of the possibilities is
that the rough design of the future exine pattern comes from the
plasma membrane with its trans-membrane integrins.
Alternatively, it could be that the trigger mechanism works “from
the other end”: the contracting MFs, attached to the plasma
membrane at many points, pull it inside, bringing about the
appearance of plasma membrane invaginations and folding of the
glycocalyx. This mechanism suggests gene-controlled contraction of
MFs (induced by ATP?). Interestingly, at late tetrad stage in Stangeria
eriopus microspore development, when the distal pole of a microspore
begins to invaginate and the form of the microspore changes from
roundish to concave, this process coincides in time and topography
with the formation of many invaginations of the plasma membrane
(Gabarayeva and Grigorjeva, 2002). These regular and deep invagina-
tions are strangely similar to those typical for middle tetrad period of
Trevesia microspores. Our current understanding of this invaginative
process in Stangeria microspores suggests the involvement of
contractile MFs in the appearance of their specific concave form.
Direct support for MFs participating in the changing of microspore
wall layer form during development comes from the necessity to use a
special relaxation solution (EGTA, MgSO
4
,NH
4
Cl which prevents
contraction of MFs —Huang and Pitelka, 1973) to keep bulges
(irregular channels) in the endexine open during fixation for TEM
preparation, otherwise the bulges would be closed (Rowley and
Skvarla, 2006).
Paxson-Sowders et al. (1997) showed that in the dex1 mutant, at
the early tetrad stage, a normal undulating pattern of the plasma
membrane is not observed. SP becomes randomly deposited on the
plasma membrane in the form of large masses, and the exine surface
lacks a normal reticulate pattern. The authors concluded that the
mutation blocked the normal invagination of the plasma membrane.
Later, Paxson-Sowders et al. (2001) isolated in this mutant the novel
plant protein, DEX1, and using freezing methods, showed not only that
plasma membrane invaginations were absent, but also that the
primexine matrix was distorted and lacked fibrillar component. The
authors predicted that the DEX1 protein was a membrane-associated
protein that may span the plasma membrane and contains calcium-
binding zones. These findings are in good accordance with our
understanding of tensegrity. If the primexine matrix is aberrant in
comparison with normal (the latter always has thinned regions
alternating with thick ones in prepatterned glycocalyx of species with
reticulate exine), then no compliant regions (which cannot resist to
inward-directed tractional forces of prestressed MFs) occur. As a result,
no invaginations would form. On the other hand, the DEX1 protein is
probably an integral protein. It is known that ECM (glycocalyx)
molecules produce cell shape changes through their binding to cell
surface integrin receptors, the latter span the plasma membrane and
interconnect with actin MFs inside the cell. Because the DEX1 protein,
and probably its distribution through the plasma membrane, is different
from that operated in wild type it results in a shifted balance and
interplay betweenthe oppositelydirected forcesof inward-directed MFs
pulling and the outward-directed resistance of ECM–glycocalyx, hence
the aberrant microspore form. If the glycocalyx (primexine matrix) is
distorted, no regular SP accumulation would be expected.
4.2.4. “Hollow”columellae
The differential contrast of the columellae seen at the end of middle
tetrad stage, with slight accent on the outer surface (Plate VI, 5),
drastically increases by the late tetrad stage, when the first
accumulation of SP takes place, so that columellae appear as “hollow”
structures in longitudinal (Plate VII, 3, 4) and cross (Plate VIII)
sections. At first sight, such images seem to be striking and unknown
in developmental palynology, but this initial impression is soon
replaced by the familiar motif of hollow structures (without quotation
229N. Gabarayeva et al. / Review of Palaeobotany and Palynology 156 (20 09) 211–232
marks). These are seen in work on the chemical degradation of mature
exines of several species (Blackmore and Claugher, 1987; Blackmore,
1990) and SEM-traced ontogeny of Tragopogon microspores (Black-
more and Barnes, 1987). The authors, using oxidation by potassium
permanganate applied after conventional acetolysis procedure,
observed in the region of ectexine a series of interconnected hollow
spaces delimited by a boundary layer. The interpretation was that this
skeletal ectexine represented a remnant of the primexine correspond-
ing to the receptive sites that first define the form of ectexinous
structural elements (Blackmore and Barnes, 1987). In immature
microspores of Echinops, the primexine consisted of a system of
hollow tubes, corresponding to columellae in the mature wall.
Blackmore (1990) has interpreted these as receptive sites of the
primexine. Our data on the effect of 4-methylmorpholine N-oxide
monohydrate (MMNO. H
2
O), a potent solvent for polysaccharides, on
mature exines of a number of species showed comparable results,
especially prominent with Betula pollen where hollow columellae
appear after the treatment (Rowley et al., 2001, Figs. 7–9). The
diameter of the columellae is large (about 250 nm) and comparable
with those of Trevesia (350–400 nm). Another destruction experi-
ment, using oxidation of mature exines with potassium permanganate
after acetolysis, demonstrated the appearance of hollow columellae in
Lavatera pollen walls and hollow alveolae in Stangeria alveolate
exines, erosion increasing with the duration of treatment (Gabarayeva
et al., 2003). Our interpretation was that the core subunits of tufts,
which didn't resist oxidation, were sites of accumulation of receptor-
independent SP (for the latter term see Rowley and Claugher, 1991).
All these findings clearly show that “hollow”columellae, observed
in Trevesia late tetrad stage, are that very ontogenetic step which is
usually missed in ontogeny studies —most probably because of its
briefness. We think that the outer parts of the primexine “hollow”
columellae correspond to the receptive sites for receptor-dependent
SP that first define the form of ectexinous structural elements (the
“boundary layer”of Blackmore and Barnes, 1987).
Thus cylindrical micelles inside composite procolumellae at
first do not accumulate SP (Plate VII, 4, arrow), then gradually do so
(Plate VIII, 1) until only narrow SP-free channels remain.
Images like those shown in Plate VII, 3, arrowhead and inset, where
a gigantic spiral twists round the procolumella, may demonstrate that
several micelles–tufts, being packed together, undergo re-distribution
of their subunits. Their cores become united, and their binder units
form one huge common spiral around the whole procolumella.
The same gigantic spirals around the columellae were observed in
Cabomba exine (Gabarayeva et al., 2003b). In this case a new arrange-
ment of united tufts–micelles would correspond to the appearance of
one cylindrical super-micelle. This suggestion receives support from
the fact that “when two or more similar micelles encounter each other
they may merge”(Ball, 1994). In such a case, what we see as a
“hollow”columella is an inverted super-micelle, with a hydrophilic
interior allowing passage of water-soluble nutrients from the tapetum
—a microchannel, and a hydrophobic surface.
4.2.5. Summarizing remarks on tetrad period
Our results and interpretations can be compared with results on
male-sterile mutants of Arabidopsis thaliana (Ariizumi et al., 2004,
2005). A novel mutant nef1 exists which completely lacks exine
(Ariizumi et al., 2004). Only a coarse, sponge-like primexine matrix
appears in this mutant microspore, predestined by very shallow
plasma membrane invaginations. There is no procolumellae and no
normal SP accumulation. Instead, the whole microspore becomes
surrounded with scale-like agglutinations of SP. Lipid analysis
revealed that the total lipid content was lower than that in the wild
type, and that the mutant has a defect in fatty acid and/or lipid
biosynthesis in the plastids of the tapetum (Ariizumi et al., 2004). All
this developmental collapse is seen in the light of our understanding
as follows: (1) lack of lipids from the early tetrad stage would
destabilise the colloidal composition in the periplasmic space,
preventing normal formation of the micellar glycocalyx and would
result in the appearance of a coarse, disordered framework; (2) the
primexine matrix (glycocalyx) would lack lipopolysaccharides —a
constituent of the glycocalyx (Rowley, 1975); (3) indistinct, shallow
and irregular invaginations of the plasma membrane (which are
probably the consequence of unbalanced tensegrity resulting from a
discord in CSK net and/or disorder in integrin–MF connection —the
more so that NEF1 protein is predicted to be an integral membrane
protein which maintains the integrity of the cell envelope) would
lead to aberrant architecture of the glycocalyx (unlike the normal
reticulate prepattern, typical for a future reticulate exine); (4) initial
disorder in the glycocalyx tuft–micelle arrangement would lead to
the absence or abnormal distribution of SP receptors, and a defect in
fatty acid and/or lipid biosynthesis could result in the lack of
receptor-dependent SP, the consequence of which would be the
absence of procolumellae at the middle tetrad stage and the
disorderly appearance of receptor-independent SP agglutinations
around microspores and in the anther loculus; (5) microspores
lacking in the normal primexine with its plasmodesmata-like
cylindrical micelles–tufts would be incapable of nutrient absorption
and therefore inviable. The other Arabidopsis male-sterile mutant,
isolated by T-DNA tagging screening, did not have a normal thick
primexine. Instead, a moderately electron-dense layer is formed
around the microspores. Later, no reticulate exine is formed, but SP
accumulates in the form of an exine-like coarse structure. The
authors (Ariizumi et al., 2005) concluded that the electron-dense
layer was an aberrant primexine which partly played its role in SP
deposition. We suggest that in this mutant the distribution of SAPs
was also disturbed or absent, resulting in spontaneous, receptor-
independent SP accumulation.
The tetrad period is, in essence, the most importantone in ectexine
development: all the main events have already occurred, and the
exine “design”is set. Further SP accumulation during the post-tetrad
period is only a quantitative increase of its mass, derived from pure
physico-chemical processes, i.e. self-assembly of colloidal emulsions.
Self-assembly also directs the formation of the second exine layer —
the endexine.
4.3. Partial conclusions
The study of pollen wall development in Trevesia burckii has
revealed the following sequence of events and their interpretations:
1. The early post-meiotic tetrad stages show mainly globular
accumulations in the periplasmic space, between callose and the
plasma membrane, at some transitive moments with addition of
tubule and branching structures, often conjugated with globules.
These are most probably spherical micelles, with the addition of
cylindrical ones, and, in a way, hybrid micelles at the transition
from spherical to cylindrical, which capture lipophilic substances of
tapetal or/and microspore origin.
2. The next structures in the periplasmic “arena”are clusters of
roundish entities with a dark core and light contrasted halo,
fragments of membranes, and a fine network —a kind of molecular
and supramolecular “soup”of spherical micelles, their remnants
and their progenitors.
3. Clusters of roundish entities are gradually replaced by radially
oriented tuft-units. These evidently correspond to clusters of
spherical micelles which self-arrange into cylindrical ones attached
to the plasma membrane. The latter becomes regularly invaginated.
This stage is a conventional junction between transitive prepara-
tory stages and the establishment of the glycocalyx (primexine
matrix), a scaffolding for exine.
4. New data obtained in studies with mutants, confirm previous
opinions on the importance of callose for the exine pattern
230 N. Gabarayeva et al. / Review of Palaeobotany and Palynology 156 (2009) 211–232
formation. A callose jacket around a microspore not only restricts a
special narrow slit (an arena for most important processes of the
tetrad period), imposing necessary constraints in the periplasmic
space and promoting the increase of glycoprotein concentration
and micelle formation, but it may behave actively in respect to the
adjacent macromolecular layer (glycocalyx), promoting structure-
forming process at the interface.
5. At middle tetrad stage the plasma membrane becomes deeply and
regularly i nvaginated, with the glycocalyx occur ring as bunches of
radially oriented cylindrical micelle–tufts on pinnacles (procolu-
mellae), capable of further growth. However, within invaginations
the glycocalyx is detached from the plasma membrane and never
undergoes further development. The microspores acquire a
typical hedgehog-like form. The supposition is that tensegrity
principles are involved in the change of the tetrad microspore
form in species with a reticulate pattern. A hypothesis for a
sequence of tensegrity events is proposed, starting with the
principle of cell pre-stress.
6. The plasma membrane itself may be a major factor in pattern
determination. If it really bears a 2-dimensional sketch of the
future pattern from the premeiotic sporophytic stage, a special
mechanism must exist to convert 2-dimensional information into
3-dimensional (ontogenetic time will be the third parameter). It
might also be that the plasma membrane's integrins, being the sites
of adhesion between exocellular matrix and elements of the
cytoskeleton, play a very important role in pattern determination
(for instance, if they bear SAPs or promote their distribution).
7. Sporopollenin accumulation starts from the outer “walls”of
procolumellae and progresses centripetally. Therefore the images
suggest “hollow”columellae. Such an initial outline of the future
ectexine pattern (boundary layer, in terms of S. Blackmore) is a
typical feature of the middle tetrad stage. This phenomenon is in
good accordance with experiment data, revealing the extreme
resistance of these sites to oxidation —most probably because they
are locations of sporopollenin polymerisation promoters
(enzymes?), the sporopollenin-acceptor (-attractor?) particles
(SAPs, by J. Rowley). Successive massive receptor-independent
sporopollenin accumulations in the post-tetrad period are less
resistant to oxidation.
4.4. General conclusions
1. This study confirms the initially suggested hypothesis (Gabarayeva
and Hemsley, 20 06; Hemsley and Gabarayeva, 2007)about
considerable participation of self-assembling micelle transitive
mesophases to construction of the framework of exine —
glycocalyx; direct genomic control is linked mainly to the location
of sporopollenin receptors through this 3-D framework.
2. Tensegrity regularities, being self-assembling, are also suggested to
be involved in the establishment of the exine, timed for the middle
tetrad stage.
Acknowledgements
This study was supported by RFBR grant No. 08-04-00498 to Nina
Gabarayeva. We thank Bruce Sampson for reading and improving
the language. We also are thankful to Peter Cinnman for the as-
sistance with the TEM Hitachi-H-600, and to Nikolay Arnautov for
the opportunity to use the material from the green-houses for this
study.
References
Adams, J.C., Watt, F.M., 1993. Regulation of development and differentiation by the
extracellular matrix. Development 117, 1183–1198.
Anger, E., Weber, M., 2006. Pollen wall formation in Arum alpinum. Ann. Bot. 97,
239–244.
Ariizumi, T., Hatakeyama, K., Hinata, K., Inatsugi, R., Nishida, I., Sato, S., Kato, T., Tabata,
S., Toriyama, K., 2004. Disruption of the novel plant protein NEF1 affects lipid
accumulation in the plastids of the tapetum and exine formation of pollen, resulting
in male sterility in Arabidopsis thaliana. Plant J. 39, 170–181.
Ariizumi, T., Hatakeyama, K., Hinata, K., Sato, S., Kato, T., Tabata, S., Toriyama, K., 2005.
The HKM gene, which is identical to the MS1 gene of Arabidopsis thaliana,is
essential for primexine formation and exine pattern formation. Sex. Plant Reprod.
18, 1 –7.
Ball, P., 1994. Designing the Molecular World. Princeton University Press, Princeton,
pp. 216–255.
Barnes, S.H., Blackmore, S., 1986. Some functional features in pollen development. In:
Blackmore, S., Ferguson, I.K. (Eds.), Pollen and Spores: Form and Function.
Academic Press, London, pp. 71–80.
Bissel, M.J., Hall, H.G., Parry, G., 1982. How does extra-cellular matrix direct gene
expression? Theor. Biol. 99, 31–68.
Blackmore, S., 1990. Sporoderm homologies and morphogenesis in land plants, with a
discussion on Echinops sphaerocephala (Compositae). Plant Syst. Evol. 5, 1–12.
Blackmore, S., Barnes, S.H., 1987. Pollen wall morphogenesis in Tragopogon porrifolius L.
(Compositae: Lactuceae) and its taxonomic significance. Rev. Palaeobot.Palynol. 52,
233–246.
Blackmore, S., Barnes, S.H., 1990. Comparative studies of mature and developing pollen
grains. In: Claugher, D. (Ed.), Scanning Electron Microscopy in Taxonomy and
Functional Morphology. Systematics Association Special, vol. 41. Clarendon Press,
Oxford, pp. 1–21.
Blackmore, S., Claugher, D., 1987. Observations an the substructural organization of
the eine in Fagus sylvatica L. (Fagaceae) and Scorzonera hispanica L. (Compositae:
Lactuceae). Rev. Palaeobot. Palynol. 53, 175–184.
Blackmore, S., Wortley, A.H., Skvarla, J.J., Rowley, J.R., 2007. Pollen wall development in
flowering plants. New Phytol. 174, 483–498.
Borsh, T., Wilde, V., 2000. Pollen variability within species, populations, and individuals,
with particular reference to Nelumbo. In: Harley, M.M., Morton, C.M., Blackmore,
S. (Eds.), Pollen and Spores: Morphology and Biology. Royal Bot. Gardens, Kew,
pp. 285–299.
Buckminster Fuller, R., 1965. Conceptuality of fundamental structures. In: Kepes, G.
(Ed.), Structure in art and in science. Braziller, New York, pp. 66–87.
Caredda, A., Devaux, P., Sangwan, R., Clément, C., 1999. Differential development of
plastids during microspore embryogenesis in barley. Protoplasma 208, 248–256.
Clément, C., Laporte, P., Audran, J.C., 1998. The loculus content and tapetum during
pollen development in Lilium. Sex. Plant Reprod. 11, 94–10 6.
Collinson, M.E., Hemsley, A.R., Taylor, W.A., 1993. Sporopollenin exhibiting colloidal
organization in spore walls. Grana, Suppl. 1, 31–39.
Dickinson, H.G., 1976a. The deposition of acetolysis-resistant polymers during the
formation of pollen. Pollen Spores 18, 321–334.
Dickinson, H.G., 1976b. Common factors in exine depositio n. In: Ferguson, I.K .,
Muller, J. (Eds.), The evolutionary significance of the exine. Academic Press,
London, pp. 67 –89.
Dickinson, H.G., 1982. The development of pollen. Rev. Cytol. Biol. Veg., Bot. 5 (1), 5–19.
Dickinson, H.G., Sheldon, J.M.,1984. A radial system of microtubules extending between
the nuclear envelope and the plasma membrane during early mail haplophase in
flowering plants. Planta 161, 86–90.
Dickinson, H.G., Willson, C., 1983. Two stages in the redifferentiation of amyloplasts in
the microspores of Lilium. Ann. Bot. 52, 803–810.
Dong, X.Y., Hong, Z.L., Sivaramakrichnan, M., Mahfouz, M., Verma, M.P.S., 2005. Callose
synthase (CalS5) is required for exine formation during microgametogenesis and
for pollen viability in Arabidopsis. Plant J. 42, 315–328.
Dunbar, A., 1973. Pollen development in the Eleocharis palustris group (Cyperaceae):
1. Ultrastructure and ontogeny. Bot. Not. 126, 197–254.
Dunbar, A., Rowley, J.R., 1984. Betula pollen development before and after dormancy:
exine and intine. Pollen Spores 26, 299–338.
El-Ghazaly, G., Jensen, W.A., 1987. Development of wheat (Triticum aestivum) pollen.
II. Histochemical differentiation of wall and Ubish bodies during development.
Am. J. Bot. 74, 1396–1418.
Elgsaeter, A., Stokker, B.T., Mikkelsen, A., Branton, D., 1986. The molecular basis of
erythrocyte shape. Science 234, 1217–1223.
Fridrichsberg, D.A., 1995. Colloidal chemistry. Publish House “Chemistry”,St.-Petersburg.
Gabarayeva, N.I., 1986. The development of the exine in Michelia fuscata (Magnoliaceae)
in connection with the changes in cytoplasmic organelles of microspores and
tapetum. Bot. Z. 71, 311–322 (In Russian, with English summary).
Gabarayeva, N.I., 1987. Ultrastructure and sporoderm development in Manglietia
tenuipes (Magnoliaceae) in the tetrad period: the establishment of primexine in
connection with functioning of cytoplasmic organelles. Bot. Z. (Leningrad) 72,
281–290 (In Russian, with English summary).
Gabarayeva, N.I., 1990. Hypothetical ways of exine structure determination. Bot. Z.
(Leningrad) 75, 1353–1362 (In Russian, with English summary).
Gabarayeva, N.I.,1991.Patterns of development in primitive angios perm pollen. In: Blackmore,
S., Barnes, S.H. (Eds.), Pollen and Spores. Clarendon Press, Oxford, pp. 257–268.
Gabarayeva, N.I., 1992. Sporoderm development in Asimina triloba (Annonaceae). I. The
developmental events before callose dissolution. Grana 31, 213–222.
Gabarayeva, N.I., 1993. Hypothetical ways of exine pattern determination. Grana 33
(suppl. 2), 54–59.
Gabarayeva, N.I., 1995. Pollen wall and tapetum development in Anaxagorea brevipes
(Annonaceae): sporoderm substructure, cytoskeleton, sporopollenin precursor
particles, and the endexine problem. Rev. Palaeobot. Palynol. 85, 123–152.
Gabarayeva, N.I., 2000. Principles and recurrent themes in sporoderm development.
In: Harley, M.M., Morton, C.M., Blackmore, S. (Eds.), Pollen and Spores: Morphology
and Biology. Royal Bot Gardens, Kew, pp. 1–17.
231N. Gabarayeva et al. / Review of Palaeobotany and Palynology 156 (2009) 211–232
Gabarayeva, N.I., Rowley, J.R., 1994. Exine development in Nymphaea colorata
(Nymphaeaceae). Nord. J. Bot. 14, 671–691.
Gabarayeva, N.I., El-Ghazaly, G., 1997. Sporoderm development in Nymphaea mexicana
(Nymphaeaceae). Plant Syst. Evol. 204, 1–19.
Gabarayeva, N.I., Grigorjeva, V.V., 2002. Exine development in Stangeria eriopus
(Stangeriaceae): ultrastructure and substructure, sporopollenin accumulation,
the equivocal character of the aperture, and stereology of microspore organelles.
Rev. Palaeobot. Palynol. 122, 185–218.
Gabarayeva, N.I., Gr igorjeva, V.V., 2003. Comparative study of the pollen wall
development in Illicium floridanum (Illiciaceae) and Schisandra chinensis (Schisan-
draceae). Taiwania 48, 147–167.
Gabarayeva, N.I., Hemsley,A .R., 2006. Merging concepts: the role of self-assembly in the
development of pollen wall structure. Rev. Palaeobot. Palynol. 138, 121–139.
Gabarayeva, N.I., Rowley, J.R., Skvarla, J.J., 1998. Exine development in Borago
(Boraginaceae). 1. Microspore tetrad period. Taiwania 43, 203–214.
Gabarayeva, N.I., Grigorjeva, V.V., Rowley, J.R., 2003. Sporoderm ontogeny in Cabomba
aquatica (Cabombaceae). Rev. Palaeobot. Palynol. 127, 147–173.
Gerasimova-Navashina, E.N., 1973. Physicochemical nature of primexine formation of
angiosperm pollen grains. In: Kovarski, A. (Ed.), Embryology of angiosperms.
Kishinev, pp. 57–70.
Griffiths, P.C., Hemsley, A.R., 2001. Raspberries and muffins —mimicking biological
pattern formation. Colloids Surf., B Biointerfaces 25, 163–170.
Goodwin, B., Kauffman, S., Murray, J.D., 1993. Is morphogenesis an intrinsically robust
process? J. Theor. Biol. 163, 135–144.
Grushvitsky, I.V., Skvortsova, N.T., Arkhangelsky, D.B., Chistyakova, L.D., 1984. The
species of the genus Trevesia (Araliaceae) in the Flora of Vietnam. Bot. Z. 69,
1019–1029 (in Russian, with English summary).
Gunning, B.E.S., Steer, M.W., 1986. Plant cell biology. An ultrastructural approach.
University College Dublin, Dublin.
Hamley, I., 2000. Introduction to soft matter. Polymers, colloids, amphiphiles and liquid
crystals. John Wiley, Chichester.
Hamm, M., Goldbeck-Wood, G., Fraaije, J.G.E.M., Zvelindowsky, A.V., 2004. Nematic-
amorphous polymer interfaces in the presence of a compatibilizer. J. Chem. Phys.
121, 4430–4440.
Harris, A.K., Wild, P., Stopak, D., 1980. Silicone rubber substrata: a new wrinkle in the
study of cell locomotion. Science 208, 177–180.
Hemsley, A.R., Gabarayeva, N.I., 2007. Exine development: the importance of looking
through a colloid chemistry “window”. Plant Syst. Evol. 263, 25–49.
Hemsley, A.R., Collinson, M.E., Brain, A.P.R., 1992. Colloidal crystal-like structure of
sporopollenin in the megaspore walls of recent Selaginella and similar fossil spores.
Bot. J. Linn. Soc. 108, 307–320.
Hemsley, A.R., Collinson, M.E., Kovach, W.L., Vincent, B., Williams, T., 1994. The role of
self-assembly in biological systems: evidence from iridescent colloidal sporopolle-
nin in Selaginella megaspore walls. Phil. Trans. R. Soc. Lond. B 345, 163–173.
Hemsley, A.R., Jenkins, P.D., Collinson, M.E., Vincent, B.,1996. Experimental modelling of
exine self-assembly. Bot. J. Linn. Soc. 121, 177–187.
Hemsley, A.R., Vincent, B., Collinson, M.E., Griffiths, P.C., 1998. Simulated self-assembly
of spore exines. Ann. Bot. 82, 105–109.
Hemsley,A.R.,Collinson,M.E.,Vicent,B.,Griffiths,P.C.,Jenkins,P.D.,2000.Self-assemblyof
colloidal units in exine development. In: Harley, M.M., Morton, C.M., Blackmore, S.
(Eds.), Pollen and Spores: Morphology and Biology. Royal Bot. Gardens, Kew, pp.
31–44.
Hemsley, A.R., Lewis, J., Griffiths, P.C., 2004. Soft and sticky development: some
underlying reasons for microarchitectural pattern convergence. Rev. Palaeobot.
Palaynol. 130, 105–119.
Heslop-Harrison, J., 1968. Tapetal origin of pollen coat substances in Lilium. New Phytol.
67, 779–786.
Heslop-Harrison, J., 1972. Pattern in plant cell walls: morphogenesis in miniature. Proc.
Royal Inst. Great Britain, vol. 45, pp. 335–351.
Hesse, M., 1985. Hemispherical surface processes of exine and orbicules in Calluna
(Ericaceae). Grana 24, 93–98.
Hesse, M., 1995. Cytology and morphogenesis of pollen and spores. Prog. Bot. 56,
33–55.
Huang, B., Pitelka, D.R., 1973. The contractile process in the Ciliate, Stentor coeruleus.I.
The role of microtubules and filaments. J. Cell Biol. 57, 704–728.
Ingber, D., 1993. Cellular tensegrity: defining new rules of biological design that govern
the cytoskeleton. J. Cell Sci. 104, 613–627.
Ingber, D.E., 2003a. Tensegrity I. Cell structure and hierarchical systems biology. J. Cell
Sci. 116, 1157–1173.
Ingber, D.E., 2003b. Tensegrity II. How structural networks influence cellular
information processing networks. J. Cell Sci. 116,1397–1408.
Ingber, D.E., Jamieson, J.D., 1985. Cells as tensegrity structures: architectural regulation
of histodifferentiation by physical forces transduced over basement membrane. In:
Andersson, L.C., Gahmberg, C.G., Ekblom, P. (Eds.), Gene expression during normal
and malignant differentiation. Academic Press, Orlando, FL, pp. 13–32.
Lipowsky, R., 1991. The conformation of membranes. Nature 349, 475–481.
Lugardon, B., 1990. Pteridophyte sp orogenesis: a survey of spore wall ontogeny and fine
structure in a polyphyletic plant group. In: Blackmore, S., Knox, R.B. (Eds.),
Microspores: Evolution and Ontogeny. Academic Press, London, pp. 95–120.
Mascarenhas, J.P., 1975. The biochemistry of angiosperm pollen development. Bot. Rev.
41, 259–314.
Mittal, K.L., Mukerjee, P.,1977. Wide world of micelles. In: Mittal, K.L. (Ed.), Micellization,
Solubilization, and Microemulsions, vol. 1,2. Plenum Press, NY, pp. 11–31.
Muñoz, C.A., Webster, B.D., Jernstedt, J.A., 1995. Spatial congruence between exine
pattern, microtubules and endomembranes in Vigna pollen. Sex. Plant Reprod. 8,
147–151.
Nishikawa, S., Zinkl, G.M., Swanson, R.J., Maruyama, D., Preuss, D., 2005. Callose (β-1,3
glucan) is essential for Arabidopsis pollen wall patterning, but not tube growth.
BMC Plant Biol. 5, 22–30.
Paxson-Sowders, D.M., Owen, H.A., Makaroff, C.A., 1997. A comparative ultrastructural
analysis of exine pattern development in wild-type Arabidopsis and a mutant
defective in pattern formation. Protoplasma 198, 53–65.
Paxson-Sowders, D.M., Dodrill, C.J., Owen, H.A., Makaroff, C.A., 2001. DEX1, a novel
plant protein, is required for exine pattern formation during pollen development in
Arabidopsis. Plant Physiol. 127, 1739–1749.
Pettitt, J.M., 1979. Ultrastructure and cytochemistry of spore wall morphogenesis. In:
Dyer, A.F. (Ed.), The experimental biology of ferns. Acad. Press, London, pp.
211–252.
Pettitt, J.M., Jermy, A.C., 1974. The surface coats on spores. Biol. J. Linn. Soc. 6, 245–257.
Robenek, H., 1980. Relationships between the endoplasmic reticulum and the plasma
membrane of protoplasts of Scimmia japonica Thunb. Biol. Zent.bl. 99, 13–23.
Rodríguez-García, M.I., Majewska-Sawka, A., 1992. Is the special callose wall of
microsporocytes an impermeable barrier? J. Exp. Bot. 43, 1659–1663 .
Rowley, J.R., 1971. Implications on the nature of sporopollenin based upon pollen
development. In: Brooks, J., Grant, P.R., Muir, M., van Gijzel, P., Shaw, G. (Eds.),
Sporopollenin. Academic Press, London, pp. 174–219.
Rowley, J.R., 1975. Lipopolysaccharides embedded within the exine of pollen grains.
In: Bailey, G.W. (Ed.), 33rd Ann. Proc. Electron Microscopy Soc. Amer. Las Vegas,
Nevada, pp. 572–573.
Rowley, J.R., 1990. The fundamental structure of the pollen exine. Plant Syst. Evol.,
Suppl. 5, 13–29.
Rowley, J.R., Flynn, J.J., 1968. Tubular fibrils and the ontogeny of the yellow water lily
pollen grain. Cell Biol. 39, 159.
Rowley, J.R., Dunbar, A., 1970. Transfer of colloid iron from sporophyte to gametophyte.
Pollen Spores 12, 305–328.
Rowley, J.R., Dahl, A.O., 1977. Pollen development in Artemisia vulgaris with special
reference to glycocalyx material. Pollen Spores 19, 169–284.
Rowley, J.R., Claugher, D., 1991. Receptor-independent sporopollenin. Bot. Acta 104,
316–323.
Rowley, J.R., Claugher, D., 1996. Structure of the exine of Epilobium angustifolium
(Onagraceae). Grana 35, 79–86.
Rowley, J.R., Rowley, J.S., 1998. Stain reversal in pollen exines. In: Narendra, M.D. (Ed.),
Current concepts in pollen-spore and biopollination research. Research Periodicals
and Book Publishing House, India, pp. 223–232.
Rowley, J.R., Skvarla, J.J., 2006. Pollen development in Epilobium (Onagraceae): late
microspore stages (a review). Rev. Palaeobot. Palynol. 140, 91–112 .
Rowley, J.R., Gabarayeva, N.I., Walles, B., 1992a. Cyclic invasion of tapetal cells into loculi
during microspore development in Nymphaea colorata (Nymphaeaceae). Amer. J.
Bot. 79, 801–808.
Rowley, J.R., Skvarla, J.J., Pettitt, J.M., 1992b. Pollen wall development in Eucommia
ulmoides (Eucommiaceae). Rev. Palaeobot. Palynol. 70, 297–323.
Rowley, J.R., Gabarayeva, N.I., Skvarla, J.J., El-Ghazaly, G., 2001. The effect of 4-
methylmorpholine N-oxide monohydrate (MMNO. H
2
O) on pollen and spore
exines. Taiwania 46, 246–273.
Rowley, J.R., Skvarla, J.J., El-Ghazaly, G., 2003. Transfer of material through the
microspore exine —from the loculus into the cytoplasm. Can. J. Bot. 81, 1070–1082.
Sattler, R.,1992. Towards a more dynamic plant morphology.Acta Biotheor. 38, 303–315.
Scott, R.J.,1994. Pollen exine —the sporopollenin enigmaand the physics of pattern.In: Scott,
R.J.,Stead, M.A.(Eds.), Molecularand Cellular Aspects of PlantReproduction.Society for
Experimental Biology Seminar Series, vol. 55. Cambridge Univ. Press, pp. 49–81.
Scriven, L.E., 1977. Equilibrium bicontinuous structures. In: Mittal, K.L. (Ed.), Micelliza-
tion, Solubilization, and Microemulsions, vol. 1,2. Plenum Press, N.-Y., pp. 548–567.
Skvarla, J.J., Rowley, J.R., 1987. Ontogeny of pollen in Poinciana (Leguminosae).
I. Development of exine template. Rev. Palaeobot. Palynol. 50, 239–311.
Sheldon, J.M., Dickinson, H.G., 1983. Determination of patterning in the pollen wall of
Lilium henryi. J. Cell Sci. 63, 191–208.
Sheldon, J.M., Dickinson, H.G., 1986. Pollen wall formation in Lilium: the effect of
chaotropic agents, and the organisation of the microtubular cytoskeleton during
pattern development. Planta 168, 1–23.
Sims, J.R., Karp, S., Ingber, D.E., 1992. Altering the cellular mechanical force balance
results in integrated changes in cell, cytoskeletal and nuclear shape. J. Cell Sci. 103,
1215–1222.
Southworth, D., Jernstedt, J.A., 1995. Pollen exine development precedes microtubule
rearrangement in Vigna unguiculata (Fabaceae): a model for pollen wall patterning.
Protoplasma 187, 79–87.
Takahashi, M., 1993. Exine initiation and substructure in pollen of Caesalpinia japonica
(Leguminosae: Caesalpinioidea). Am. J. Bot. 80, 192–197.
Tryon, A.F., Lugardon, B.,1978. Wall structure and mineral content in Selaginella spores.
Pollen Spores 20, 315–340.
van Blaaderen, A., 2006. Colloids get complex. Nature 439, 545–546.
vanUffelen,G.A.,1991.Thecontrolofsporewallformation.In:Blackmore,S.,Barnes,S.H.
(Eds.), Pollen and spores: Patterns of Diversification. Clarendon Press, Oxford, pp.
89–102.
Wang, N., Butler, J.P., Ingber, D.E., 1993. Mechanotransduction across the cell surface and
through the cytoskeleton. Science 260, 1124–1127 .
Worrall, D., Hird, D.L., Hodge, R., Paul, W., Draper, J., Scott, R., 1992. Premature
dissolution of the microsporocyte callose wall causes male sterility in transgenic
tobacco. Plant Cell 4, 759–771.
Wyatt, S.E., Carpita, N.C., 1993. The plant cytoskeleton–cell-wall continuum. Trends Cell
Biol. 3, 413–417.
Xu, Ting-Yu, 1982. Angiosperm pollen flora of tropic and subtropic China. Inst. Bot.
South China, pp. 17–428.
232 N. Gabarayeva et al. / Review of Palaeobotany and Palynology 156 (2009) 211–232
