I. Primexine development in Passiflora racemosa Brot.: overlooked aspects of development

Developmental stages during the tetrad period were examined in detail by transmission electron microscopy with an emphasis on substructure. Our purpose was to find out whether the sequence of sporoderm developmental events provides additional evidence for our recent hypothesis on the underlying cause of exine ontogeny as a sequence of self-assembling micellar mesophases initiated by genomically given physicochemical parameters. Osmiophilic globules encrusting the surface of postmeiotic microspores and tapetal cells are temporary prepattern units which come first. The second prepattern structures are highly ordered bundles of microfilaments and microtubules which determine the position of microspore surface invaginations and clusters of the glycocalyx inside them. The first glycocalyx units are microgranules which during the middle tetrad stage rearrange into radially oriented rod-like units. The latter form lens-like clusters of the glycocalyx-1, located inside the invaginations. These clusters predestine the position of the future luminae in the exine reticulum. The second glycocalyx layer is laid down as a continuous layer over the whole microspore surface and has similar substructure, that is radial rods. Glycocalyx-2 is a framework for procolumellae which appear at the late tetrad stage. Therefore, the sequence of substructural units in the primexine is: globules, microgranules, rod-like units, and layers of radially oriented rods tightly packed in the periplasmic space. This sequence corresponds to the first three mesophases of self-assembling micelles: spherical micelles, cylindrical micelles, and layers of hexagonally packed cylindrical micelles (middle mesophase). We observed the same sequence in other species during primexine development, and the findings of this study provide new evidence for our hypothesis.     

Pollen wall ontogeny in <Emphasis Type="Italic">Polemonium caeruleum</Emphasis> (Polemoniaceae) and suggested underlying mechanisms of development

By a detailed ontogenetic study of Polemonium caeruleum pollen, tracing each stage of development at high TEM resolution, we aim to understand the establishment of the pollen wall and to unravel the mechanisms underlying sporoderm development. The main steps of exine ontogeny in Polemonium caeruleum, observed in the microspore periplasmic space, are spherical units, gradually transforming into columns, then to rod-like units (procolumellae), the appearance of the initial tectum, growth of columellae in height and tectum in thickness and initial sporopollenin accumulation on them, the appearance of the endexine lamellae and of dark-contrasted particles on the tectum, the appearance of a sponge-like layer and of the intine in aperture sites, the appearance of the foot layer on the base of the sponge-like layer and of spinules on the tectum, and massive sporopollenin accumulation. This sequence of developmental events fits well to the sequence of self-assembling micellar mesophases. This gives (together with earlier findings and experimental exine simulations) strong evidence that genome and self-assembly probably share control of exine formation. It is highly probable that self-assembly is an intrinsic instrument of evolution.     

Tapetal Cell Changes and Sporoderm Development in Rhigozum trichotomum (Burch.)

Rhigozum trichotomum is a perrenial woody shrub which is indigenousto the arid regions of southern Africa. Primexine developmentis initiated while the microspores are still enclosed by callose.This is followed by the appearance of probacula which give riseto the tectum, bacula and nexine. At the time of callose dissolution,the exine pattern is well established and intine developmenthas been initiated. During the tetrad stage, the protoplastsof the tapetal cells exhibit shrinkage while conspicuous stacksof rough endoplasmic reticulum become evident in their cytoplasm.These stacks produce numerous vesicles which are associatedwith lipid globules and which migrate to the tapetal/locularwall where, it is suggested, they give rise to the pro-orbicules.The pro-orbicules become coated with an osmiophilic substance,probably sporopollenin, and are released into the thecal fluidto become intimately bound to the exine, Here they are strippedof the osmiophilic layers which appear to be incorporated intothe sporoderm. Rhigozum trichotomum (Burch.), sporoderm, pollen wall, exine, orbicules, pro-orbicules, sporopollenin, tapetum     

Sporoderm development in Acer tataricum (Aceraceae): an interpretation

For the first time, the developmental events in the course of complicated exine structure establishment have been traced in detail with transmission electron microscope in the representative of Acer. A new look at unfolding events is suggested using the knowledge of a boundary field, colloid science. Our purpose was to find out whether the sequence of sporoderm developmental events represents, in essence, the sequence of self-assembling micellar mesophases, initiated by genomically given physicochemical parameters and induced by surfactant glycoproteins at increasing concentration. Indeed, the first units observed in the periplasmic space are globular ones (=spherical micelles) which become arranged into rod-like units (=cylindrical micelles). Then, twisted clusters of rodlets form a layer of procolumellae (middle micellar mesophase). The tectum emerges as an untwisting and merging of distal ends of procolumellae (distal untwist of micelle clusters). In the end of tetrad period, when a hydrophilic–hydrophobic switch occurs in the periplasmic space, the contrast reversal of the columellae corresponds to the change of normal micelles to reverse ones. The initiation of the foot layer and the endexine lamellae, with their typical central “white lines”, corresponds to the next—“neat”—mesophase, with its typical central gaps between layers. Aperture sites during development show all the main micellar mesophases and their transitional forms. The data received have supported our previous hypothesis.     

Exine development in Nymphaea colorata (Nymphaeaceae)

We show a sequence of developmental events in microspores and tapetal cells in Nymphaea colorata based upon transmission and scanning electron microscopic observations. There are parallel cytoplasmic processes and surface coatings in microspores and tapetal cells. Uptake is indicated by the passage of lanthanum as a tracer from anther locule into the microspore cytoplasm and by the condition of the cytoplasmic surface of microspores. The callose envelope is not a barrier to transfer of lanthanum. During formation of the proexine glycocalyx tiny spiral elements, components of the exine substructural units, were oriented in different directions in the surface coating of microspores and tapetum. Lipoidal globules are associated with the spiral elements. After the uniform proexine stage, three regions of different exine structure and their gradations become differentiated in the sporoderm: 1) a proximal region with thick tectum and foot layer, thin columellae and a compact layer of lamellated endexine; 2) a distal pole region with separately disposed endexine lamellae; and 3) an equatorial encircling-sulcate aperture region which consists of infratectal layer, foot layer, and endexine lamellae. Based upon the presence of structurally comparable surface coats in microspores and tapetal cells, experimental uptake of lanthanum nitrate, and the co-ordinated processes in tapetum and microspores, we conclude that there is probably a reciprocal controlling influence between the microspores and the tapetum and other sporophytic tissues.     

Microspore and tapetal development in Echinodorus cordifolìus (Alismaceae)

In the microspore tetrad period the exine begins as rods that originate from the plasma membrane. These rods are exine units that on further development become columellae as well as part of the tectum, foot layer and “transitory endexine”. The primexine matrix is very thin in the future sites of the pores. At these sites the plasma membrane and its surface coating (glycocalyx) are without exine units and adjacent to the callose envelope. The exine around the aperture margin is characterized by units of reduced height. After the exine units and primexine matrix have become ca 0.2 μm in height a fibrillar zone forms under the aperture margin. It is the exine units around the aperture that are templates for exine processes on apertures of mature pollen. Oblique sections of the early exine show that the tectum consists of the distal portions of close-packed exine units. The exine enlarges in the free microspore period but initially its substructure (tectum, columellae, foot layer and transitory endexine) is not homogeneous and unit structures are visible until after the vacuolate microspore period. There are indications of a commissural line/plane (junction plane) which separates the foot layer from the endexine during early development. Our observations of development in Echinodorus pollen extend a growing number of reports of “transitory endexines” in monocot pollen. The exine unit-structures become 0.2 μm or more in diameter and many columellae are composed of only one exine unit. Spinules become exceptionally tall, many protruding ca 0.7 μm above the level of the tectum as units only ca 0.1 μm in diameter. The outer portion of the tectum fills in around spinules and by maturity they are microechinate with their bases spread out to ca 1 μm or more. Unit structures can be seen with SEM in mature pollen following oxidation by plasma ashing and in the tapetum these units are arranged both radially, as in spinules, and parallel with the tapetal surfaces. There are clear indications of such an arrangement of units in untreated fresh pollen. Units comprising the basal part of the exine are not completely fused by sporopollenin accumulated during development. This would seem to be a characteristic feature, based on published work, of the alismacean pollen. Our use of a tracer shows, however, that there is considerable space within or between exine structure of mature Echinodorus pollen. Based upon the ca 0.1 μm size of exine-units formed early in development and exine components seen after oxidative treatment it seems that the early (primary) accumulated sporopollenin has greater resistance to oxidation than sporopollenin added, secondarily, around and between units later in development. Both primarily and secondarily accumulated sporopollenin are resistant to acetolysis but published work indicates that acetolysis alters exine material. At the microspore tetrad time and until the vacuolate stages tapetal cells are arranged as in secretory tapetums. During early microspore stages there are orbicules at the inner surface of tapetal cells. At free microspore period tapetal cells greatly elongate into the loculus and surround the microspores. By the end of the microspore vacuolate period tapetal cells release their cellular contents and microspores are for a time enveloped by tapetal organelles and translocation material.     

Exine ontogeny in Borago officinalis pollen

The primexine matrix is finely granulo-fibrillar up to callose digestion; it becomes distinctly fibrillar at the free microspore stage. The columellae and the tectum are initiated at the middle tetrad stage, the foot layer and the endexine are initiated when the callose wall digestion begins. The columellae are initiated by the deposition of spiral elements around a clear central zone. This hollow aspect of columella disappears when thickening. The foot layer and the endexine are built by the expansion of plasmalemma derived components. The foot layer appears first at the poles, then at the interapertural levels and at last at the apertures while the endexine appears first at the mesoapertures, then it spreads laterally towards the interapertural levels and, at last, at the poles. The gemmae are formed at the free microspore stage over all the tectum. The thickening of the exine takes place essentially during the free microspore stage and continues during the vacuolate microspore one. Apertures are entirely formed before the complete digestion of the callose wall. The ectoapertures are determined by the lacking of the columellae; the sites of the pericolpal cavities and the mesoapertures result from the plasmalemma retraction even before the setting up of the foot layer and the endexine by which they will be delimited respectively afterwards. The endoapertures are determined by the lacking of compact endexine at their level, and merge into a continuous equatorial belt.     

Pollen morphology of Zehneria s.l. (Cucurbitaceae)

A growing body of experimental data obtained from sporoderm ontogenetic studies led to the appearance of the ‘micellar’ hypothesis. The hypothesis is that the sequence of sporoderm developmental events represents the sequence of self-assembling micellar mesophases, initiated by genomically given physico-chemical parameters, which are then picked up by physico-chemical self-assembly. However, besides morphological evidence, the best proof of this hypothesis would be an experimental modelling of sporoderm-like patterns. The main idea of this study is to remove the influence of the genome, selecting substances and their concentrations for simulations to replace it, and then to trace what ‘pure’ self-assembly is capable of constructing. Our aim in this study was to simulate mainly young structures in sporoderm development, i.e. the glycocalyx and the primexine. Several polysaccharide gels (as a callose substitute) and surfactants (as glycocalyx and sporopollenin monomer substitutes) were mixed at different concentrations and combinations, thermally set and left to condense. A number of patterns were obtained in colloidal solutions in the course of condensation, simulating structures at different stages of exine development. Their structures were observed and analysed with transmission electron microscopy (TEM). Our first experiments on the modelling of biological patterns in vitro have shown encouraging results.     

Microsporogenesis in Pinus sylvestris. VI. Exine and tapetal development during the tetrad period

Coverage is of microspore tetrad period from end of cytokinesis to introduction of endexine in Pinus sylvestris. The ectexine of aperture, cap zone and sacci and the endexine are initiated while microspores are in the tetrad condition and enveloped in callose. Ectexine patterning including considerable expansion of sacci develops prior to the initiation of the endexine. Alveoli, sacci and alveoli within sacci are initiated by cytoplasmic invaginations which are sites of uptake of cell surface coat (glycocalyx) along with nutrients bound to the glycocalyx. Applications of tracers show that glycocalyx elements bind to cations and transport them to the cytoplasm. From the beginning of exine formation these invaginations are largest in the regions of future sacci and very small in the aperture. As growth progresses cytoplasm surrounding invaginations partially retracts, but callose contact is retained. Thus, these invaginations become callose covered hemispheroids (alveoli) that are “open” to the cell surface proximally and covered by callose distally but only partially so at the sides of the “cup‐shaped” alveoli. Until introduction of the endexine part of the alveolar‐sides are made up of cytoplasmic protrusions which contact the callose protrusions, even across sacci expanded more than 3 μm. Glycocalyx elements become aligned on the inner surface of the callosic alveoli and are sites for sporopollenin accumulation. The template for endexine components consists of glycocalyx elements that become aligned near the plasma membrane. Our observations indicate that uptake from the loculus to the microspore cytoplasm changes after introduction of the endexine. Henceforth, uptake is assisted by the endexine, as shown by tracers. Tapetal cells undergo two periods of hyperactivity during the period covered. Hyperactivity took place at the beginning of uptake by microspores and during endexine formation. The extra tapetal lamellation and its tapetal markers begin to exhibit the intense staining, after endexine initiation.     

POLLEN WALL AND TAPETAL ORBICULAR WALL DEVELOPMENT IN SORGHUM BICOLOR (GRAMINEAE)

Pollen wall development in Sorghum bicolor is morphologically and temporally paralleled by the formation of a prominent orbicular wall on the inner tangential surface of the tapetum. In the late tetrad stage, a thin, nearly uniform primexine forms around each microspore (except at the pore site) beneath the intact callose; concurrently, small spherical bodies (pro-orbicules) appear between the undulate tapetal plasmalemma and the disappearing tapetal primary wall. Within the primexine, differentially staining loci appear, which only develop into young bacula as the callose disappears. Thus, microspore walls are devoid of a visible exine pattern when released from tetrads. Afterwards, sporopollenin accumulates simultaneously on the primexine and bacula, forming the exine, and on the pro-orbicules, forming orbicules. Channels develop in the tectum and nexine, and both layers thicken to complete the microspore exine. Channeled sporopollenin also accumulates on the orbicules. A prominent sporopollenin reticulum interconnects the individual orbicules to produce an orbicular wall; this wall persists even after the tapetal protoplasts degenerate and after anthesis. While the pollen grains become engorged with reserves, a thick intine, containing conspicuous cytoplasmic channels, forms beneath the exine. Fibrous material collects beneath the orbicular wall. The parallel development and morphological similarities between the tapetal and pollen walls are discussed.     

II. Exine development in Passiflora racemosa Brot.: post-tetrad period. Overlooked aspects of development

Developmental stages during the post-tetrad period are examined in detail with TEM and SEM, with emphasis upon substructure. Our purpose was to find out whether the sequence of sporoderm developmental events gives additional evidence for our recent hypothesis on the underlying cause of exine ontogeny as a sequence of self-assembling micellar mesophases, initiated by genomically given physico-chemical parameters. Four different layers of the endexine are developed in the post-tetrad period. The first one is a layer of white line centered lamellae which appear as a demarcation line between ect- and endexine. The second layer is sponge-like and consists of “roots” of columellae and a layer between them. The third layer consists of basally disposed radially elongated granules which appear in the aperture sites only. The fourth layer emerges in interapertural sites only and is formed as stacks of uneven lamellae. Therefore, the sequence of substructural units in primexine is the next: white-lined lamellae, a layer of honeycombed substructure, globule-to-rod-like granules, stacks of wavy lamellae. These sequences correspond to the next four mesophases of self-assembling micelles: neat (=laminate) micelles, high-concentrated emulsion of sponge-like (=foam-like) substructure, spherical-to-cylindrical micelles, and laminate micelles with fenestrated laminae. Reiteration of the micellar mesophases, participating in endexine development, is observed during the post-tetrad period.     

A new look at sporoderm ontogeny in Persea americana and the hidden side of development

Background and Aims

The phenomenon of self-assembly, widespread in both the living and the non-living world, is a key mechanism in sporoderm pattern formation. Observations in developmental palynology appear in a new light if they are regarded as aspects of a sequence of micellar colloidal mesophases at genomically controlled initial parameters. The exine of Persea is reduced to ornamentaion (spines and gemmae with underlying skin-like ectexine); there is no endexine. Development of Persea exine was analysed based on the idea that ornamentation of pollen occurs largely by self-assembly.

Methods

Flower buds were collected from trees grown in greenhouses over 11 years in order to examine all the main developmental stages, including the very short tetrad period. After fixing, sections were examined using transmission electron microscopy.

Key Results and Conclusions

The locations of future spines are determined by lipid droplets in invaginations of the microspore plasma membrane. The addition of new sporopollenin monomers into these invaginations leads to the appearance of chimeric polymersomes, which, after splitting into two individual assemblies, give rise to both liquid-crystal conical ‘skeletons’ of spines and spherical micelles. After autopolymerization of sporopollenin, spines emerge around their skeletons, nested into clusters of globules. These clusters and single globules between spines appear on a base of spherical micelles. The intine also develops on the base of micellar mesophases. Colloidal chemistry helps to provide a more general understanding of the processes and explains recurrent features of pollen walls from remote taxa.     

Exine formation in the pollinium ofDendrobium

Summary The position of the callose wall is related to the position of the primexine matrix that forms around the peripheral tetrads during microspore development of the compound unit, the pollinium. We report a combined freeze-fracture and freeze-substitution study of the events associated with early exine development. Stage one of exine development is deposition of protosporopollenin that is probably synthesised by the microspore and secreted to the primexine matrix where it is polymerised. Enzymes for the polymerisation of the protosporopollenin may be synthesised by the microspores and then transported, via the endoplasmic reticulum, to the plasma membrane. Stage two of exine development follows callose dissolution and deposition of tapetally derived sporopollenin. Hence exine form and exine deposition inDendrobium appear to be the result of intimate cooperation between the microspore, the plasma membrane, the callose and the tapetum.     

Microsporogenesis in Taxus baccata L.: The Formation of the Tetrad and Development of the Microspores

Following meiosis II in Taxus microsporangia a small proportionof the tetrads regularly degenerated. Despite frequent inequalityin the frequency of ribosomes between the spores of a tetrad,partial degeneration within a tetrad was never observed. Theinitial wall of the young spores was found to resemble the wallof the mother cell in containing a fibrillar layer, and thetwo walls may possess similar isolating properties. The symmetryof the tetrad was regularly iso-bilateral. The formation ofthe sporoderm began as the spores were released into the loculusby the rapid dissolution of the wall of the mother cell. Osmiophilicdroplets emerged from the spore protoplast and entered the wall.The fibrillar layer ceased to be recognizable and the dropletscoalesced to form an outer layer on which up to six sporopolleninlamellae, probably of tapetal origin, were deposited. The accretionof a single layer of sporopollenin droplets, in no recognizablepattern, gave rise to the outer verrucose part of the exine.Cytochemical tests showed that the tapetum was rich in acidphosphatases from the beginning of meiosis. Towards the endof its degeneration the tapetum intruded into the loculus andcould therefore be regarded as partly invasive. Taxus baccata, microsporogenesis, tetrad symmetry, sporoderm     

ONTOGENETIC AND EVOLUTIONARY IMPLICATIONS OF A NEOTENOUS EXINE IN TAPEINOCHILOS (ZINGIBERALES: COSTACEAE) POLLEN

Tapeinochilos pollen, like that of most angiosperms, is spared by the standard acetolysis treatment because the sporoderm is impregnated with sporopollenin. This genus and its allies in the Costaceae are the only taxa in the eight families of Zingiberales that have acetolysis-resistant pollen. The sporoderm in most of the order is characterized by exine reduced to a wispy coating or layer with delicately anchored spinules and a highly elaborated intine. Ultrastructural studies on the pollen of Tapeinochilos reveal a pattern of wall development that is significantly different from the generalized angiosperm type; namely, there are no columellae, nor is there any significant accretion of sporopollenin following the dissolution of callose and release of microspores. The primexine is composed of rodlets which build up solidly between apertures and become packed into layers to form a thick, stratified exinous covering. No secondary exine develops during the free spore period and the juvenile primexine persists as the protective coat on the mature pollen grain. This pattern of pollen development is viewed as an example of neoteny in which a juvenile or immature character is retained in adulthood.     

Pollen wall characters with emphasis upon applicability

The message for exine pattern resides ultimately in the genome, yet the information for the initial form exists in the cytoplasm or with the plasma membrane and its glycocalyx. Subsequent wall development is likely to be the result of an interplay between the genome, the cytoplasm, and the intralocular environment. The exine consists of units derived from the plasma membrane glycocalyx and enveloped in the exinous polymer sporopollenin. Growth of the exine in accomodation to cell surface expansion is modeled as involving a doubling in diameter of the units, separation of components, and incorportation of new units within the nexine but not the tectum. If the tectum is thick and does not become disjunct, its restraint upon cellular expansion may result in the crushing of bacules. Both the final shape and ornamentation of the exine may be influenced by cytological processes like oncoid plugs that limit the effect of protoplast expansion to nonapertural regions or globules in the exine arcade that can cause distention of the tectum and rupturing of bacules. Subunits of exinous units can be seen in distinctive patterns at the outer surface of the exine, in the arcade of the exine, and prior to intine formation at the inner surface of the nexine.     

Development of the pollen wall in Tsuga canadensis (Pinaceae)

In discussions of exine structural types, Tsuga is often mentioned as an exception, since no infratectal layer is present in the ektexine. The present investigation documents the formation of this pollen wall type at the ultrastructural level in T. canadensis . All layers of the exine are formed during the tetrad period, when the microspores are surrounded by a callose wall. The outer layer (ektexine) is elaborated on a fibrillar microspore surface coat, while the inner layer (endexine) is elaborated on lamellated structures. The deposition of the pretectum is followed by the appearance of endexine lamellae. In the initial stages, the two layers—pretectum and endexine—appear to be separated from each other only by a dense microspore surface coat. As additional wall materials are deposited, the tectal elements become convoluted and come to rest, in places, on the now recognizable footlayer. Upon release from the tetrad, intine formation begins and continuous accumulation of sporopollenin leads to an increase in ektexine thickness. The mature pollen wall of Tsuga canadensis , with a convoluted tectum resting directly on the footlayer, is characteristic of the genus.     

Studies on the Ontogeny of the Pollinium of Goodyera Procera

Using light, transmission and scanning electron microscopy, the development of the pollinium of Goodyera procera (Ker-Gawler) Hooker. was investigated. At the early stage, sporogenous cells inside the microsporangium were seen grouping together into small aggregates each containing few cells. After the aggregates have formed the sporogenous cells inside the aggregates (which could now be called massulae) divide to form numerous pollen mother cells. Later, the pollen mother cells undergo meiosis to form tetrads. The pattern of formation of the exine of tetrads varies according to the location of the tetrads inside the micro- sporangium. Those tetrads that are situated near the outer region of the massulae can form: exine with well developed tectum, bacula and foot layer; and the sequence of events leading to the formation of this type of well developed exine is as follows the original wall and the cyto- plasmic channels associated with the wall become surrounded by a thick layer of callose thus isolating the wall from the plasmalemma. Near the plasmalemma a layer of primexine containing callose and cellulose begins to form. Later, the primexine develops into exine and between the exine and plasmalemma a layer of intine is laid down. Similar type of exine with well developed tectum, bacula and foot layer, is also present in tetrads facing the tapetum. But in this case the original wall of the tedtrad is not retained but undergoes dissolution and in its place a new exine formed. The pattern of formation of exine in the region between tetrads is even more different. Here the original wall also undergoes dissolution but instead of forming a proper exine it only forms a thin foot layer with bulges at places. The pattern of formation of the exine in the cells inside the tetrad is even more different. Here the original wall of the cells only undergoes partial dissolution. The loose fibrils of the partially dissolved wall then become mixed with the callose layer surrounding the cell. Inside this wall-fibril/callose mixture thin sheets of exine appear, but these thin sheets of exine do not develop further into tectum or bacula. In Goodyera a quite substantial amount of callose is retained in the regions between massulae and tetrads, and we believe that it is this callose which is holding the massulae and tetrads together to form pollinium.     

Development and cytochemistry of pollen and tapetum in Mitriostigma axillare (Rubiaceae)

The development of pollen grains and tapetum in Mitriostigma axillare (Rubiaceae) was studied from anther primordium to dehiscence. Anthers were freeze-cracked and studied with SEM. Embedded anthers were sectioned and studied with LM and TEM. Cytochemistry was performed in order to distinguish the different layers of the sporoderm and to determine its chemical nature at different development stages. The pollen grains remained as tetrads by partial fusion of the exine, probably because of reduced callose septa during the stage of microspore tetrads within callose envelopes. Characteristic features of the sporoderm were an irregular foot layer, an endexine composed of amalgamated granules, a transient granular-fibrous layer beneath the endexine, and a thin intine. During maturation of the exine, the endexine became chemically different from the ectexine. All layers of the sporoderm were reduced in thickness due to stretching during the engorgement of the pollen grains prior to dehiscence. The pollen grains were colpoidorate with a reticulate to microreticulate tectum covered with a scanty surface coating. The mature pollen grains were binucleate and contained a lot of starch grains. Thick intineous onci protruded through the apertures and formed papillae. About 50% of the microspores were aborted. The tapetum was of secretory type, probably with cycles of hyperactivity and protrusions of the cells into the locular cavity. No syncytium was formed and there were neither orbicules nor tapetal membrane.     

Exine and tapetum development in Nymphaea capensis (Nymphaeaceae): a comparative study

The developmental events in microspore envelope and cytoplasm and in tapetum from premeiosis until late tetrad stage were studied in Nymphaea capensis. The exceptional feature of microspore development in this species is that post-meiosis cytokinesis is retarded until the late tetrad stage. Thus, the entire development of the exine becomes completed during the tetrad stage. As a consequence of the retarded cytokinesis, the proximal portion of the forming exine lags behind the distal one during the major part of the tetrad period, but eventually the proximal part of the exine overtakes the distal part in development. The significance of this retardation is discussed. This sequence of events differs sharply from corresponding sporoderm development in other Nymphaea species. Another important topic is the microspore surface activities during exine development. The surface coatings-glycocalyx-are very similar in microspores and in tapetum cells, but their functions are completely different; the roots for this difference are discussed. A noteworthy feature of the developing microspores is the presence of gigantic, deeply cup-like mitochondria; this property is also characteristic of the microspore cytoplasm of N colorata and N. mexicana. A functional significance of these organelles and their adaptive role is discussed.