|Home | About | Journals | Submit | Contact Us | Français|
In flowering plants, microsporogenesis is accompanied by various types of cytoplasmic partitioning (cytokinesis). Patterns of male cytokinesis are suspected to play a role in the diversity of aperture patterns found in pollen grains of angiosperms. The relationships between intersporal wall formation, tetrad shape and pollen aperture pattern ontogeny are studied.
A comparative analysis of meiosis and aperture distribution was performed within tetrads in two triporate eudicot species with contrasting aperture arrangements within their tetrads [Epilobium roseum (Onagraceae) and Paranomus reflexus (Proteaceae)].
Intersporal wall formation is a two-step process in both species. Cytokinesis is first achieved by the formation of naked centripetal cell plates. These naked cell plates are then covered by additional thick, localized callose deposits that differ in location between the two species. Apertures are finally formed in areas in which additional callose is deposited on the cell plates. The recorded variation in tetrad shape is correlated with variations in aperture pattern, demonstrating the role of cell partitioning in aperture pattern ontogeny.
To disperse their male gametes, flowering plants produce pollen grains. At maturity, each pollen grain comprises two or three cells enclosed within a highly elaborated wall composed mostly of sporopollenin, one of the most resistant compounds produced by a living organism. The pollen wall protects the male gametophyte against UV radiation, pathogenic attack and desiccation during pollen dispersal. In most species the wall is interrupted by apertures, which are areas generally characterized by a thinning of the external wall layer (exine) and a thickening of the internal wall layer (intine). Apertures perform several important functions in securing pollination success. They prevent pollen wall breakage by permitting water loss (dehydration) and water uptake (rehydration) on the stigma. They are further associated with signal exchanges during pollen rehydratation and finally permit germination of the pollen tube (Edlund et al., 2004). The early evolution of seed plants was characterized by the acquisition of a set of key innovations associated with the male gametophyte. There was a reversal of microspore polarity (a shift from proximal aperture development relative to the tetrad to distal aperture development), a loss of sperm mobility and co-option of the pollen tube to achieve siphonogamy (Rudall and Bateman, 2007). A second shift in aperture position and aperture number occurred within angiosperms, with the single distal polar aperture replaced by three equatorial apertures (tricolpate pollen) at the base of the eudicot clade. This apertural change is thought to have been important in the subsequent radiation of the clade by facilitating contact between at least one aperture and the stigmatic surface (Furness and Rudall, 2004).
The main features of pollen wall morphology are determined during male meiosis (microsporogenesis), the first step in pollen development. A thick callose wall is first formed around the microsporocytes, which will persist until after the completion of meiosis. The next step in the meiotic process sees the formation of a tetrad composed of four microspores (Fig. 1A–D). During the tetrad stage, microspores produce a primexine template of the future pollen wall. Apertures are usually visible before the release of microspores from the tetrad. The aperture pattern is defined by the shape, number and distribution of apertures on the pollen grain surface within the tetrad (Walker and Doyle, 1975).
Species that produce pollen grains with three equatorial apertures have one of two arrangements of the apertures within the tetrad. In Fischer's arrangement (Fischer, 1890) apertures form pairs at six points in the tetrad, while in Garside's arrangement (Garside, 1946) they form groups of three at four points in the tetrad. In both cases the apertures meet in places where the microspores remain in contact until just before their release from the tetrad (Fig. 1E). Wodehouse (1935) first proposed that the meeting points of apertures within the tetrad are determined by the pattern of meiotic cytokinesis. More precisely, he proposed that the spatial information determining aperture location is provided by the last contact points between the cytoplasms of the future microspores at the end of cytokinesis. This leads to grouped apertures at the contact points between microspores (Wodehouse, 1935). Understanding aperture pattern ontogeny and evolution in angiosperms thus requires scrutiny of the relationships between cytokinesis, tetrad formation and aperture patterns.
Cytokinesis following meiosis is achieved by the formation of intersporal callose walls, which leads to the formation of a tetrad. Considerable variation in the pattern of microsporogenesis has been described among the angiosperms (reviewed in Bandhari, 1984; Blackmore and Crane, 1988; Furness et al., 2002). Ressayre et al. (2002a) recorded this variation and proposed a developmental model that considers these variations when explaining and predicting aperture patterns. The variation affects (a) the timing of nuclear divisions relative to cytoplasmic divisions, (b) the orientation of the meiotic axes (between the first and second nuclear division axes and/or between the second division axes) and (c) the way in which callose intersporal walls are formed during cytokinesis. The first two types of variation (successive versus simultaneous cytokinesis) and the orientation of meiotic axes (spatial variation) affect the shape of the tetrad. The third type of variation (pattern of callose deposition to form walls between microspores) determines the completion points of cytokinesis (Ressayre et al., 2002a). Data showing the correlation between developmental variation and aperture pattern diversity are accumulating (Ressayre, 2001; Ressayre et al., 2002b, 2003, 2005; Furness and Rudall, 2004; Nadot et al., 2006; Blackmore et al., 2007).
This relationship between cytokinesis and aperture formation and its applicability across angiosperms can be tested by comparing the pattern of microsporogenesis in species with different aperture patterns. This was done by Ressayre et al. (2005) in four species with contrasting aperture patterns (polar versus equatorial apertures and Fischer's versus Garside's arrangement of apertures within the tetrad). In these species, the observed pattern of microsporogenesis was the one expected to produce the observed distribution of apertures within the tetrads (Ressayre et al., 2002a). As such it provided strong arguments supporting the implication of post-meiotic cytokinesis in the control of aperture pattern determination.
However, the relationship between aperture location and tetrad shape so far remains unexplored from a developmental point of view. The present paper aims to fill this void. Here, microsporogenesis and aperture distribution within tetrads were studied in two eudicot species with intraspecific variation in tetrad shape and aperture type, and with contrasting arrangements of apertures within tetrahedral tetrads between the two species. Epilobium roseum (Onagraceae) displays Fischer's arrangement, while Paranomus reflexus (Proteaceae) displays Garside's arrangement. Onagraceae and Proteaceae differ in their arrangement of pores (Fischer's versus Garside's arrangement), but are superficially similar in having more or less triangular pollen grains with pores situated at the angles (angulaperturate). This renders them ideal as a system for testing hypotheses on the mechanisms that lead to the different aperture positions. This system allows both hypotheses (Wodehouse, 1935; Ressayre et al., 2002a) to be tested. Clear evidence of a correlation between tetrad shape, intersporal wall formation and aperture pattern is presented.
Seeds of Epilobium roseum were provided by the Botanical Garden of Caen (France) (accession 1058F6). The seeds were grown in a greenhouse. Flowers of Paranomus reflexus (accession 1116/83) were sampled in Kirstenbosch Botanical Garden (South Africa) in July 2002.
Fresh flower buds were collected at various developmental stages. Several flower buds per individual and several stamens per bud were sampled and studied for each developmental stage. The anthers were dissected out, immediately squashed and mounted in anilin blue to which 15 % glycerol had been added, following a modified protocol of Arens (1949). With this method, callose becomes fluorescent when UV illuminated (DAPI or FITC filters). In Paranomus reflexus, fully formed tetrads and pollen grains were stained with congo red (Stainier et al., 1967), which emphasizes the contrast between apertural zones and the rest of the wall. In Epilobium roseum mature pollen grains are dispersed in tetrads. The tetrads were mounted in anilin blue and observed using a DAPI filter. In both species the progression of intersporal callose wall formation, the deposition of additional callose, the resulting tetrad shape and the arrangement of apertures within the tetrad were recorded.
Epilobium roseum and Paranomus reflexus differ in terms of how they complete cytokinesis, i.e. in the distribution of additional thick callose deposited onto the cell plates. This difference is linked to the difference in aperture pattern observed between the two species.
Cytokinesis is simultaneous and occurs through the formation of centripetal cell plates (Fig. 2A, B). Three discrete tetrad shapes were observed: tetrahedral (Fig. 2C–E), tetragonal (Fig. 2F–H) and rhomboidal (Fig. 2I–L). The tetrahedral shape is by far the most common. When cytokinesis is almost complete, the last connections between the cytoplasms of the future microspores (last contact points) are positioned at the centre of the cleavage wall (arrows in Fig. 2B, I). In all tetrads, cytokinesis is completed through a large deposition of additional callose on the cell plates (Fig. 2C, D, F, G, J, K). In tetrahedral tetrads, callose depositions mostly accumulate in the centre of the tetrad and at the intersection of the cell plates (Fig. 2C, D,a–c). In tetragonal tetrads, these callose deposits form two large masses (a square on one side, and a cross on the other) located from the intersections between the four cell plates to the outer wall of the tetrad (Fig. 2F, G,a–c). In rhomboidal tetrads, the callose deposits also exhibit an asymmetric distribution, forming two large masses (a rectangle on one side, and a cross on the other) located from the intersections between the cell plates to the outer wall of the tetrad (Fig. 2J, K,a–c). In all tetrads, additional callose is last deposited in places close to the observed last points of contact; these places are at a slightly higher focal level (Fig. 2, arrows in C, F, J). At a later developmental stage, the last points of callose deposition coincide with positions in which aperture pairs are seen (Fig. 2E,a, H,a, L,a). The apertures meet in pairs, conforming to Fischer's tetrad arrangement. Depending on both the shape of the tetrad and orientation of the pollen grains therein, different aperture patterns are observed in the mature pollen grains (pollen is dispersed in a tetrad). Tetrahedral tetrads are composed of four pollen grains with three pores each (Fig. 2E,a,b), tetragonal tetrads are made up of four pollen grains with two pores each (Fig. 2H,a,b) and rhomboidal tetrads of two central pollen grains with three pores each and two external pollen grains with two pores each (Fig. 2L,a,b).
Cytokinesis is simultaneous and cell plates are formed centripetally (Fig. 3A). Most of the tetrads are tetrahedral (Fig. 3A, B, F–I), although tetragonal tetrads were occasionally observed (Fig. 3D, E). A few tetrads were difficult to characterize clearly (Fig. 3C, J), and could not be assigned to any of the three discrete types of tetrads usually recorded (tetragonal, tetrahedral or rhomboidal). They represent intermediates between tetragonal and tetrahedral tetrads with unequal cleavage walls (Fig. 3C), from here on referred to as ‘intermediate’ tetrads. In all tetrads, intersporal wall formation is completed by large additional callose deposits on each cell plate (Fig. 3D, F–H, J,a–c). In tetrahedral tetrads, the last points where callose is deposited correspond to the places where cell plates intersect in groups of three (Fig. 3G, H). These callose deposits are always thicker towards the outer wall of the tetrad, becoming thinner towards the centre of the tetrad, irrespective of tetrad shape (tetragonal tetrads, Fig. 3D; tetrahedral tetrads, Fig. 3F,a,b, G, H,a,b; intermediate tetrad, Fig. 3J,a–c). Mature pollen grains present two to four pores (Fig. 3K–M). Diporate pollen grains are ovoid with the two pores distributed at each end of the longest axis, triporate pollen grains are triangular and tetraporate pollen grains are quadrangular. Triporate pollen grains are much more common than diporate grains, while tetraporate grains are very rare (Fig. 3N). Although aperture arrangement could not be observed while the microspores were still in tetrads, it could easily be inferred from the shape of the pollen grains. Apertures of diporate, triporate and tetraporate pollen grains are joined, in two opposite groups of four in tetragonal tetrads (Fig. 3E), in four groups of three in tetrahedral tetrads (Fig. 3I, thus following Garside's arrangement), and in three groups of four or three in intermediate tetrads (Fig. 3J,a–c).
Epilobium roseum (Onagraceae, rosids) and Paranomus reflexus (Proteaceae, basal eudicots) both simultaneously produce several different pollen morphs that differ in their aperture number (within the same anther and even in some cases within the same tetrads). In E. roseum, pollen grains have two or three apertures, while in P. reflexus pollen grains can have two, three or four apertures. The formation of microspores with different aperture numbers is directly associated with variations in the shape of the tetrads in both species. To study the processes underlying the formation of these different morphs, the different developmental events during microsporogenesis that are known to influence aperture pattern ontogeny were studied. Although cell-plate formation occurs centripetally in both species, the completion of cytokinesis differs between the two in terms of the distribution of additional thick callose deposited onto the cell plates. Observations on the pattern of intersporal wall formation and the shape of mature pollen tetrads in E. roseum, and of intersporal wall formation and post-meiotic tetrads in P. reflexus clearly show that the location of aperture groups within tetrads of both species coincide with the places where callose was deposited last. This supports the hypothesis that aperture pattern ontogeny is linked to post-meiotic cytokinesis.
Within any given tetrad, the positions of final callose deposition coincide with the positions in which apertures will be formed. Intersporal wall formation is a two step process in both species. Isolated callose cell plates are first formed centripetally, after which wall formation is completed through the deposition of additional thick callose deposits. Callose is not evenly deposited onto the cell plates, but is concentrated at specific positions that differ between the two species. In E. roseum, callose deposits are thick at cell plate intersections in the centre of the tetrad and at the intersection of the cleavage walls, becoming thinner towards the periphery of the tetrad. In P. reflexus callose deposits are the thickest in the middle of each cell plate and in the centre of the tetrad, decreasing in thickness towards the intersections of the cleavage walls. In both species the deposition of additional callose deposits relative to the cell plates does not depend on the number of cell plates (whatever the number of cleavage planes, the deposition within a cleavage plane is the same). But, different tetrad shapes correspond to different numbers of cleavage planes. As a result, tetrad shape affects the number and distribution of the places where callose is deposited last.
Although apertures were not observable within post-meiotic tetrads, the aperture distribution within tetrads could be deduced in both species. In Epilobium, mature pollen grains are dispersed in tetrads, enabling the description of aperture distribution within tetrads even if aperture formation occurs late in development (Rowley and Skvarla, 2004). In this species, tetragonal tetrads include four diporate pollen grains, tetrahedral tetrads include four triporate pollen grains, and rhomboidal tetrads include two diporate and two triporate pollen grains. In all of these tetrads apertures are joined in pairs, following Fischer's rule. Mature rhomboidal and tetragonal tetrads are not formed by post-meiotic alterations such as tetrahedral tetrad break-down or shape modification. Indeed these two tetrad types, although rare (around 1 % of the tetrads within an anther), can be observed at different stages of development in post-meiotic tetrads still enclosed within callose. In pollen mature tetrads of all shapes, apertures meet precisely where callose is deposited last, namely close to the last point of contact between the cytoplasms of the future microspores at the end of cytokinesis. In Paranomus, the aperture type can be deduced from the shape of the mature pollen grains, despite not being visible while microspores are still assembled in tetrads. As expected in the Proteaceae, the apertures of this species are distributed according to Garside's rule (Garside, 1946; Blackmore and Barnes, 1995). The triangular-shaped triporate pollen grains are produced in tetrahedral tetrads (Fig. 3I,a–c), while diporate pollen grains are produced by planar tetragonal tetrads (Fig. 3Ea). Both triporate and diporate pollen grains are also produced in intermediate tetrads (Fig. 3J,a–c). As in E. roseum, the aperture arrangements in tetrads in P. reflexus coincide with the positions where additional callose was deposited last. In both species, the observations collected at all different stages converge to indicate that aperture pattern ontogeny is linked to the completion of the intersporal wall formation in association with tetrad shape.
In 1935, Wodehouse hypothesized that apertures are formed at the last points of contact of the cytoplasm of the resultant microspores at the end of cytokinesis (further developed by Blackmore and Crane, 1988, 1998; reviewed by Blackmore et al., 2007). Ressayre et al. (2002a) developed a model that, in some cases, predicts intra-individual variation in aperture number in plants with simultaneous cytokinesis. This model was based on a schematic description of cytokinesis and is insufficient to accommodate the wide range of variation in microsporogenesis in angiosperms. Results from the two species studied here indicate two further shortcomings in this model. Firstly, although the different tetrad shapes observed in E. roseum can be categorized into three discrete types, this was not the case in P. reflexus. Besides the ‘normal’ tetrahedral and rhomboidal tetrads, no true rhomboidal planar tetrads were observed. Instead, intermediate tetrads that are not perfectly planar were observed (Fig. 3C). Such a level of continuous variation in tetrad shape must be taken into account to fully interpret and describe the consequences of tetrad shape. Previous studies focused on monocot species already pointed to this need. Many monocot species produce irregular tetrahedral tetrads with cleavage walls of different sizes, leading to continuous variation in tetrad shape (Penet et al., 2005; Ressayre et al., 2005; Nadot et al., 2006; Sannier et al., 2006).
The second shortcoming of the model of Ressayre et al. (2002a) is the over-simplification of cytokinesis. Classically, intersporal wall formation during microsporogenesis in eudicots was described as happening through infurrowing (Farr, 1916; Wodehouse, 1935; Horner and Lersten, 1971; Albertsen and Palmer, 1979; Longly and Waterkeyn, 1979a; Blackmore and Barnes, 1988; Brown and Lemmon, 1988; Ressayre et al., 2002b). Infurrowing entails the synchronization of cell-plate formation and additional callose deposition leading to a highly recognisable pattern with future microspores separated by callose deposits that are thick at the periphery of the tetrad and decrease sharply in thickness towards the cytoplasmic regions that are still connected (see Blackmore et al., 2007, fig. 3, plate f). Such a synchronization was believed to be the rule in all core eudicots, and was indeed found in various families (e.g. Solanaceae, Brassicaceae and Asteraceae) studied to date (Wodehouse, 1935; Longly and Waterkeyn, 1979a, b; Ressayre et al., 2002b). Results in similar studies on species in the Onagraceae, Euphorbiaceae, Ranunculaceae and Proteaceae have, however, revealed that this synchronization can be broken (Blackmore and Barnes, 1995; Rowley and Skvarla, 2004; Ressayre et al., 2005; Albert et al., 2009). When isolated cell plates are first produced and subsequently embedded in callose as observed in the two species studied here, the cytoplasm of the separating microspores is separated by thin walls of callose of constant thickness (see, for example, Rowley and Skvarla, 2004, plate 1). It appears then that cytokinesis is not exclusively achieved through infurrowing in higher eudicots either.
When infurrowing takes place, the last cytoplasmic junctions between the developing microspores coincide with last sites of callose deposition. This correlation can be broken when there is a desynchronization of cell wall formation. Recent detailed observations of microsporogenesis are therefore challenging Wodehouse's last contact point hypothesis (Wodehouse, 1935). Desynchronization between cell-plate formation and additional callose deposition have made it possible to see that the last additional callose deposits – and not the actual last connection points between the cytoplasms – coincide with the positions in which apertures are formed in several different species from different families. These include: Proteaceae (Paranomus reflexus, this study; Protea lepidocarpodendron, Ressayre et al., 2005), Onagraceae (Epilobium roseum, this study), and several monocots [Pontederiaceae (Pontederia cordata, Ressayre et al., 2001), Asphodelaceae (Asphodelus albus, Ressayre et al., 2005), Hemerocallidaceae (Phormium tenax, Ressayre et al., 2005), Bromeliaceae, (Tillandsia leiboldiana, Albert et al., 2010)].
By comparing two eudicots that (a) produce pollen with the same number of apertures but that differ in terms of aperture position within tetrads and (b) exhibit intra-individual variation in aperture number, this study provides new insights into the mechanisms that determine aperture number and location on pollen grain surfaces. At the intra-individual level, pollen aperture number varies with tetrad shape, confirming a long-standing view that aperture determination is linked to cytokinesis. At the intra-specific level, comparison of aperture positions within tetrads indicate that the positions of apertures are determined by the way additional callose is deposited once the cytoplasms are severed. This study shows that both Fischer's arrangement of Epilobium roseum and Garside's arrangement of Paranomus reflexus are produced by a similar mechanism (aperture position is correlated to the last point of callose deposition) and can lead to pollen with three apertures. As the number of apertures per pollen grain is under selection (Dajoz et al., 1991), pollen grains with the same aperture number should not differ in their reproductive abilities. This suggests that although aperture number is selected, aperture arrangement within tetrad is neutral, and may explain why triaperturate pollen has independently appeared recurrently in the evolution of angiosperms either in Fischer's or Garside's arrangements.
We are grateful to L. Saunois and F. Braud for their technical assistance in the greenhouse. We thank the Kirstenbosch Botanical Garden (South Africa) and the Parc Botanique de Launay (France) for providing the plant material. This work was funded by a cooperation grant between France and South Africa no. 20 2001-1081. We conducted this research in spite of the research policies of the French government.