HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Full Text (PDF) Submit a response Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Services Email this article to a friend Similar articles in this journal Similar articles in ISI Web of Science Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via ISI Web of Science (6) Citing Articles via Google Scholar Google Scholar Articles by Tsou, C.-H. Articles by Fu, Y.-L. Search for Related Content PubMed Articles by Tsou, C.-H. Articles by Fu, Y.-L. Agricola Articles by Tsou, C.-H. Articles by Fu, Y.-L.

Tetrad pollen formation in Annona (Annonaceae): proexine formation andbinding mechanism¹

Institute of Botany, Academia Sinica, Taipei, Taiwan 115, Republic of China

Received for publication July 6, 2001. Accepted for publication October 11, 2001.

	ABSTRACT

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

 
Meiotic tetrads of Annona glabra and A. montana build up a well-developed proexine (protectum, probaculum, and pronexine) at the proximal side but only a thin pronexine at the distal side during the tetrad stage. The callosic envelope is only partially digested by the end of tetrad stage. The remaining, undigested part is composed of the intersporal mass and thin peripheral layers, and the latter is conjunct with the distal pronexine of the microspore. In this remaining callosic structure celluloses are also present. Later on, due to the continuous slow decomposition of this callose-cellulose structure and microspore expansion, microspores break up the callose-cellulose envelope. Because all the four microspores are bound together by the callose-cellulose structure, they move out of the chamber in rotation. Eventually the thin pronexine is pulled toward the center of the tetrad and the well-developed proexine becomes the distal wall. These descriptions of the partial digestion of callosic envelope, the transformation from a callose-cellulose structure to the binding system of tetrad pollen, and microspore rotation in Annona are unusual in the angiosperms.

Key Words: Annona • Annonaceae • callosic envelope • microspore rotation • pollen • proexine • tetrad

	INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

 
Compound pollen grains in the form of dyad, triad, tetrad, polyad, or pollinium occur in more than 56 families of angiosperms (Knox and McConchie, 1986 ). Two common binding systems in the permanent tetrads are recognized in about 20 families that have been examined (Knox and McConchie, 1986 ): the crosswall cohesion mechanism refers to the presence of exine bridges between the grains; ontogenetically, cytoplasmic channels are always developed between the meiotic microspores and then proexine/exine is accumulated along these channels forming the exine bridges. Normally the exine bridge is composed of both endexine and ectexine. Examples include Acacia (Kenrick and Knox, 1979 ; Knox and McConchie, 1986 ), Calluna (Dahl and Rowley, 1991 ), Drosera (Takahashi and Sohma, 1982 ), Hedycarya (Sampson, 1977 ), Onagraceae (Skvarla, Raven, and Praglowski, 1975 ), and Winteraceae (Sampson, 1981 ), etc. The simple cohesion mechanism, on the other hand, refers to the binding of adjacent microspores simply at the tectal/bacular level, without the presence of the exine bridges, as in Asimina (Waha, 1987 ; Gabarayeva, 1992, 1993a ), Elaeocharis (Dunbar, 1973 ), Leschenaultia (Knox and Friederich, 1974 ), Pyrola (Takahashi and Sohma, 1980, 1982 ), and Typha (Skvarla and Larson, 1963 ; Takahashi and Sohma, 1984 ).

On the basis of the available ontogenetical studies, in all the cases the four microspores remain in their positions throughout the development and the physical basis of forming the tetrad pollen can be attributable to either of the two above-mentioned mechanisms. Blackmore and Crane (1988) have suggested that "the cohesion of microspores in permanent tetrads to be a direct result of reduced callose synthesis (at the intersporal region)," but their suggestion has not received much attention.

Among the 56 families with compound pollen, Annonaceae, the largest family among primitive angiosperms, is one of the several families producing various pollen forms, including monads, tetrads, and polyads composed of 8, 16, or 32 grains (Walker, 1971a ). Among its approximately 120 genera (Fries, 1959 ), 40 produce compound pollen (Walker, 1971a, b ; Le Thomas, 1980, 1981 ); however, the ontogeny of compound pollen has been carried out in a single species, Asimina triloba (Waha, 1987 ; Gabarayeva, 1992, 1993a ).

The present paper reports on tetrad pollen formation in two Annona species (A. glabra and A. montana). Attention is given, in this first part of our study, to tetrad development through rotation of the microspores.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

 
Floral buds of Annona glabra and A. montana were collected from plants cultivated in Taiwan Agricultural Experimental Station in Chiayi, Taiwan. The buds were dissected into four pieces and fixed with 2% paraformaldehyde (PFA) + 2% glutaraldehyde (GA) in 0.1 mol/L phosphate buffer on ice overnight.

For transmission electron microscopy (TEM) studies
Materials were rinsed with phosphate buffer, postfixed with either (1) 1% OsO₄ in water or (2) 1% OsO₄ in buffer for 4–10 h, rinsed with buffer again, and then dehydrated through acetone series and embedded in Spurr. Thin sections were obtained and observed under a Philips (Philips, Eindhoven, Netherlands) CM-100 TEM.

For light microscopy (LM) studies
Materials were rinsed with buffer, dehydrated through t-butyl-alcohol series, embedded with paraplast, and sectioned to 7–12 µm thickness with a rotary microtome. Sections were affixed to the microscopic slides, air dried, and stored in boxes. Later on, these microscopic slides with anther sections were deparaffined with xylen, rinsed with absolute EtOH, and transferred through the alcohol series to 50% EtOH. A part of these deparaffined slides were stained with 0.1% calcofluor white for detecting the celluloses or with 0.5% aniline blue for detecting the calloses. They were examined and photographed with a Zeiss Axioplan-based fluorescent microscope. Most deparaffined slides, otherwise, were stained with Safranin O-Fast green, mounted as permanent slides, and examined under transmitted light with a Zeiss (Zeiss, Oberkochen, Germany) Axioplan microscope.

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

 
Both Annona glabra and A. montana were studied in detail from microspore mother cell (MMC) stage to free spore stage by using the TEM and LM techniques. No obvious differences were found between these two species with regard to the process of proexine formation and the binding mechanism of forming tetrad pollen, except for some minor, quantitative differences. For example, A. glabra has thinner proexine and exine, and its microspore size is smaller, etc. In order to avoid redundance, A. montana is here provided with TEM pictures (Figs. 1–40) and A. glabra LM pictures (Figs. 41–49). The description below is mainly based on A. montana but applicable to A. glabra as well except for width and height measurements.

View larger version (207K): [in this window] [in a new window]   Figs. 1–5. Annona montana. Microspore mother cells during and right after meiosis. 1. A microspore mother cell is undergoing the first meiotic division; the division is proceded by furrowing (arrows). 2. Longitudinal section of an anther. The septa are composed of tapetal cells (T). Two microspocytes are undergoing the second meiotic division. The second division plane is parallel to the anther axis (arrows). 3–4. During proexine initiation, double-membraned structures (arrows) and some unknown materials are deposited between the callosic envelope and plasmalemma. 5. Reticulate associations (arrows) are found between the callosic envelope and plasmalemma. C, callosic envelope; MW, microspore mother cell wall; N, nucleus; T, tapetum; V, vacuole. Bars in Figs. 1–5 = 10 µm, 30 µm, 1 µm, 200 nm, and 2 µm, respectively

View larger version (120K): [in this window] [in a new window]   Figs. 41–49. Annona glabra. Anthers were sectioned to 7–12 µm in thickness, stained, and viewed under a fluorescent microscope (Figs. 43, 47, and 49 ) or under a transmitted light microscope (the remaining figures). 41–42. Tetrad before (Fig. 41 ) and during (Fig. 42 ) callose digestion. The callosic envelope is stained green and the proexine dark brown. The prosexine is well developed at the proximal side (arrowheads). During callose digestion, a condensed callose layer remains at the distal surface of the microspore (arrows). 43. Anther before callose digestion. When stained with aniline blue, the thick callosic envelope shows white fluorescence. The prosexine now shows faint orange autofluorescence. 44–46. Microspore rotation. When the callosic envelope is largely digested, thin callosic tethers are still retained and conjugated with the microspore at the thin-walled side (arrows). The four microspores are held together by this callosic binding system and eventually they all rotate about 180° (Fig. 46 ). 47. Anther stained with calcofluor white to detect the celluloses. The white fluorescence (arrow) in the center of the four microspores as white lines and spots indicates the presence of celluloses in the callosic binding system. 48, 49. At the early "free microspore" stage, the callosic binding system (arrows) holds the four microspores as one tetrad. The callosic binding system turns green when stained with fast green and viewed under a transmitted light microscope as in Fig. 48 or exhibits white fluorescence when stained with aniline blue and viewed under a fluorescent microscope as in Fig. 49 . C, callosic envelope; T, tapetum. Bars in Figs. 41–49 = 40 µm, 45 µm, 40 µm, 45 µm, 45 µm, 45 µm, 50 µm, 60 µm, and 60 µm, respectively

Meiosis
The MMC is about 45 µm in diameter in the beginning of meiosis, with a thin layer of the callosic mother envelope. The cytokinesis is proceeded by furrowing (Fig. 1) and the two meiotic divisions are normally successive, resulting in a tetragonal tetrad (Fig. 2). Lipid bodies, mitochondria, plastids, vesicles, and especially vacuoles of various sizes are fairly common in the dividing cells whereas starch grains are rarely observed (Fig. 1). The four microspores originating from the two dyads are arranged in two pairs. Normally, each microspore is more or less hemispherical in cross section, with the flat, proximal side facing the second meiotic division plane. No cytoplasmic channels were found between the meiotic microspores.

Tetrad stage
Proexine initiation
The sign of proexine initiation appears at the transitional region between the proximal and the distal sides, where double-membrane-bound structures and some unknown materials are deposited in the space between the callosic envelope and plasmalemma (i.e., the interspace) (Figs. 3, 4). Sometimes, reticulate associations between the callosic envelope and the cytoplasm are also seen in this space (Fig. 5). Shortly, the membrane-bound structures disintegrate and a thin glycocalyx layer develops in the interspace over the entire proximal surface. At the same time, dark-stained sporopollenin acceptor particles (SAPs), which are about 40–60 nm wide, are seen on the top of this glycocalyx layer against the callosic envelope (Fig. 6). Later on, SAPs also occur within the glycocalyx layer, appearing in short stacks (Figs. 7, 8). The glycocalyx layer is more or less lamellate at this early stage (Figs. 6–8), and vesicles are found outside the plasmalemma. At the distal side of microspore, there is no deposition of glycocalyx layer nor SAPs (Figs. 9, 10). In samples postfixed with 1% OsO₄ in water, the distal plasmalemma is strongly undulate, with many small vesicles and invaginations along the plasmalemma (Fig. 9), which suggests that active endocytosis has taken place at this side. In those postfixed with 1% OsO₄ in buffer, the callosic envelope is much condensed, and outgrowths are seen extending toward the microspore cytoplasm forming the callosic protrusions (Fig. 10). Starting from this stage on, starch grains and lipid bodies increase prominently (Figs. 7, 9, 10).

View larger version (197K): [in this window] [in a new window]   Figs. 6–10. Annona montana. Stage of prosexine initiation of the proximal (Figs. 6–8 ) and distal sides (Figs. 9–10 ) of microspore. 6. Sporopollenin acceptor particles (SAPs) (arrows) are deposited first on the top of the glycocalyx layer (*). 7–8. SAPs (arrows) start to appear within the glycocalyx layer. 9. The distal plasmalemma is undulate, and small invaginations and vesicles (*) are common. Sample was postfixed with 1% OsO4 in water. The callosic envelope is not stained. 10. Sample was postfixed with 1% OsO4 in buffer. The callosic envelope is condensed and shows a callosic protrusion (arrow). A, autophagic vesicles; C, callosic envelope; D, dictyosome; L, lipid body; M, mitochondria; P, plastid; V, vacuole. Bars in Figs. 6–10 = 500 nm, 500 nm, 500 nm, 1 µm, and 1 µm, respectively

View larger version (170K): [in this window] [in a new window]   Figs. 11–17. Annona montana. Early phase of prosexine formation. 11–15. SAPs are distributed on the top of the glycocalyx network and as radial groups in the network. In the latter, SAPs are arranged in radial rows, leaving a central microchannel (arrows). Dictyosome vesicles are commonly found outside the plasmalemma (*). The prosexine is developed at the proximal side and diminishes abruptly toward the distal side (arrowhead) as shown in Fig. 14 . 16. SAPs start to fuse, but the microchannel of each radial group is still retained (arrows). 17. The distal side of the microspore, with no glycocalyx network nor prosexine. Note the rich cellular organelles. C, callosic envelope; D, dictyosome; L, lipid body; M, mitochondria; P, plastid. Bars in Figs. 11–17 = 500 nm, 500 nm, 400 nm, 1 µm, 1 µm, 500 nm, and 2 µm, respectively

Later on, the enlargement of SAPs through the continuing sporopollenin polymerization is noticeable. When the glycocalyx network thickens to 250 nm or so, the bacular-like probacula can be recognized due to the fusion among those SAPs in the radial groups, but nevertheless a central microchannel is retained (Fig. 16). The fusion of the SAPs on the top layer also proceeds rapidly. Thus a consolidated protectum becomes more and more distinct (Figs. 18, 20). At sites of probacula formation, the meshes of the glycocalyx network do not terminate at the surface of probacula, they travel through the probacula (Fig. 21). When the network is around 300 nm in height, the protectum and probaculum are well shaped. They are about 200 and 300 nm in height, respectively (Figs. 21, 22); dictyosome vesicles are still commonly found along the plasmalemma (Fig. 20), and the central microchannel of each probaculum is still distinct, with the diameter of 20 nm or so (Fig. 22). During this stage, no newly produced SAPs are observed within the network; if there are any of them, they must be deposited within the preexisting sites of probacula or the protectal layer. Newly initiated probacula are not found either; it seems that the timing of probaculum initiation in Annona is restricted to an early and short period of prosexine formation.

View larger version (216K): [in this window] [in a new window]   Figs. 18–22. Annona montana. Middle phase of prosexine formation. 18. Cross section of a tetrad showing the well-developed prosexine at the proximal side of each microspore. The arrows indicate the direction of the second division plane. 19. Distal side of a microspore, with no prosexine nor glycocalyx layer. The plasmalemma is undulate, with many vesicles along it. The starch grains are much consumed (*). 20. SAPs are highly fused into the protectum and probacula. No free SAP is present, and dictyosome vesicles are commonly deposited outside the plasmalemma (*). 21. The glycocalyx meshes (*) are traveling through the probacula. 22. Cross section of prosexine, showing the microchannels still remaining inside the probacula (arrows); glycocalyx meshes (**) are intermixed with the probacula. C, callosic envelope; MW, mother cell wall; N, nucleus; P, plastid; Pb, probaculum; Pt, protectum. Bars in Figs. 18–22 = 10 µm, 2 µm, 500 nm, 400 nm, and 500 nm, respectively

Pronexine formation
When the prosexine has increased to 500–600 nm in height, the pronexine starts to develop. Plasmalemma invaginations take place within the small, interprobacular regions at the proximal side and the transitional region toward the distal side (Figs. 23–25). Within those invaginated areas glycocalyx matrix is loosely organized (Figs. 23, 24). At this time, the probacula are tightly connected with the plasmalemma. Later on, a thin profoot layer is initiated from the plasmalemma first at the interbacular areas, connecting the bases of the probacula at the margin (Fig. 26), and then the formation of profoot layer beneath the probaculum (Fig. 27) and at the distal surface follows. When the profoot layer beneath the probacula is formed, the probacula, which attach to the profoot layer, become detached from the plasmalemma (Fig. 28), and a space between the plasmalemma and profoot layer is created.

View larger version (177K): [in this window] [in a new window]   Figs. 23–28. Annona montanan. Pronexine initiation. 23–24. The profoot layer is initiated at both the proximal and distal sides (arrows). Interprobacular spaces are invaginated and filled with glycocalyx materials (*). 25. Longitudinal section of a microspore, showing the well-developed prosexine at the proximal side and a thin and segmented pronexine at the distal side. The distal callosic envelope shows stratifications and many callosic protrusions (arrows). 26–27. The thin profoot layer is connected to the bases of the probacula at the periphery (arrows). The probacula are still well connected with the plasmalemma (arrowheads). Glycocalyces are loosely organized in the invaginated areas (*). 28. When the profoot layer beneath the probacula is formed, it starts to detach from the plasmalemma/cytoplasm (arrows). C, callosic envelope; L, lipid body; N, nucleus; P, plastid. Bars in Figs. 23–28 = 1 µm, 2 µm, 5 µm, 1 µm, 1 µm, and 1 µm, respectively

View larger version (172K): [in this window] [in a new window]   Figs. 29–34. Annona montanan. Late tetrad and early callose digestion stages. 29. A microspore showing well-developed proexine on one side and thin pronexine on the other side. The callosic envelope is somewhat digested. 30. The thin distal pronexine is wavy and discontinuous (arrows). Starch grains in the plastids are highly consumed. 31. Higher magnification of Fig. 29 , showing the protectum, probacula, and thin pronexine. White lines (arrows) are visible in the poendexine, which extends down to the lower glycocalyx network. 32. Higher magnification of the proendexine at the distal side, the white line (arrow) is discernible. 33–34. Lower magnifications of the cross section of a tetrad under scanning electron microscopy (Fig. 33 ) and transmission electron microscopy (Fig. 34 ). The thicker prosexine is developed at the proximal side (*) and the thin pronexine the distal. The mother cell wall is still present. C, callosic envelope; MW, mother cell wall; N, nucleus; PM, plasmalemma; V, vacuole. Bars in Figs. 29–34 = 5 µm, 400 nm, 10 µm, 1 µm, 500 nm, and 10 µm, respectively

Digestion of the callosic envelope: the composition of the callosic envelope and the binding mechanism of tetrad pollen
In this study, the TEM images of the callosic envelope prepared with different postfixatives show very different morphologies. In samples postfixed with (1) 1% OsO₄ in water, the callosic envelope is hardly stained and shows wrinkles (Figs. 6–9, 11–21, 23, 24, 29–32, 34). Not much information regarding its internal structure could be deduced. In samples postfixed with (2) 1% OsO₄ in buffer, the callosic envelope is better stained and shows stratification. It is a three-layered structure over the distal side (Figs. 3–5, 10, 25, 35, 36). The outermost one-half to two-thirds is darker and porous; inside this layer is a narrow and gray middle zone, with numerous parallel filamentous substructures. The innermost zone is very thin, as dark and amorphous as the outer zone, but without pores (Figs. 25, 35, 36). On the proximal side, the intersporal callose mass is dark and nonporous. Filamentous components are also seen running parallel to the long axis (Fig. 25).

By the time of callose digestion, the mother cell wall has disintegrated; the digestion of the callosic envelope starts from the periphery (Figs. 36, 42). Later on, the digestion is much slower. The majority of the intersporal callose mass and a thin peripheral layer surrounding the distal wall of each microspore remain undigested (Figs. 37, 42), and the thin callosic layer is tightly conjugated with the distal, wavy pronexine of the microspore (Fig. 37). Due to further slow decomposition of the calloses and the continuing microspore expansion, microspores sooner or later break the callosic envelope, usually at the junction between the distal and the proximal sides (Figs. 38, 39). Because the distal pronexine of each microspore is intimately associated with the thin peripheral callose layer (Figs. 38, 39, 44, 45), the pronexine of each microspore is gradually pulled toward the center of the tetrad, whereas the proximal, thick proexine turns out, becoming the distal wall (Figs. 38, 39, 44–46). Therefore, while the microspores are moving out of the callose chamber, they undergo an approximately 180° rotation, which results in the reversal of the proximal and distal walls of each microspore. The remaining callose structure, when tested by calcoflour white, shows distinct cellulosic bands in the center (Fig. 47). The callose-cellulose structure is further reduced and modified to a binding system in the center of the tetrad, including a central core and short peripheral tethers (Figs. 40, 48, 49). The tethers bind with the proximal nexines of the four microspores to hold them together as one tetrad. The microspores are never released completely.

View larger version (166K): [in this window] [in a new window]   Figs. 35–40. Annona montana. Stages of during and after callose digestion. 35. The callosic envelope shows three layers (1, 2, and 3). A callosic protrusion is also shown here. The sample was postfixed with 1% Osmium in buffer. 36. During callose digestion, the outer callose layer is thread-like. The pronexine is evident (*) and is associated with the callosic protrusion (arrowheads). 37–39. The thin pronexine (*) at the distal side is wavy and conjugated with the remaining thin callose layer (arrowheads). Microspores usually break the callose structure at the transitional region (arrows) and start to rotate as shown in Figs. 38 and 39 . A cell plate is still visible in Fig. 38 . 40. Microspores have already rotated 180°, with the thick prosexine toward the distal side. The callosic binding system normally has a central collum and four peripheral tethers. C, callosic envelope; CB, callosic binding system; CP, cell plate; MW, mother cell wall. Bars in Figs. 35–40 = 1 µm, 1 µm, 4 µm, 10 µm, 10 µm, and 10 µm, respectively

Pollen development
Once the microspores have come out of the callose chamber, they expand rapidly and soon fill up the anther chamber. The tapetum is compressed and gradually consumed completely. The entire process of pollen development is prolonged, the pollen tetrads enlarge to about 150 µm in diameter in Annona montana, and its distal exine thickens to about 7 µm. During this process, sporopollenins are also deposited onto the callose-cellulose binding system, which assists the persistency of this structure. When the pollen tetrads have enlarged greatly, the proximal nexines of the four grains adhere (Figs. 47, 48). Such adhesion may also help the physical binding of the tetrad during late pollen development.

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

 
A series of unusual phenomena take place in Annona glabra and A. montana during the transition from meiotic tetrad stage to free spore stage, which have not been reported in the Annonaceae or elsewhere in the angiosperms. It is worthwhile to note that all these features are correlated with the poorly understood callosic envelope. The first characteristic is the partial digestion of the callosic envelope (Figs. 39–46). The remnant callose structure as demonstrated by the calcofluor white test contains celluloses as well (Fig. 43), part of which may originate from the two meiotic cell plates that are still visible during callose digestion (Fig. 38). Pectocellulosic compounds have been shown to occur in the tetrad callosic envelope, but in the outer region only, which are probably derived from the remaining mother cell wall (Bhandari, 1981 ; Majewska-Sawka and Rodriguez-Garcia, 1999 ). Partial callose digestion in Borago has been mentioned previously. In this genus the mother callosic layer is totally digested by the end of tetrad stage, and the special callosic layer is still well retained (Gabarayeva, Rowley, and Skvarla, 1998 ), but the latter will be degraded completely at the beginning of free spore stage (Rowley, Skvarla, and Gabarayeva, 1999 ). The callose digestion in Borago proceeds in two distinct phases, but after all, the digestion is complete; such a pattern is different from having a persistent callose-cellulose structure in Annona. Nevertheless, in both cases the callose wall exhibits at least two or three distinctive layers under conventional staining procedures, which indicates the chemical heterogeneity of the callosic envelope. The processes determining the two-phase digestion in Borago and the partial digestion in Annona are undescribed, but the delayed secretion or lack of suitable enzymes must have played a role.

The second novelty of Annona glabra and A. montana is the tight conjugation between the callose-cellulose envelope and the distal, wavy pronexine of the microspore since late tetrad stage (Figs. 37–39, 44, 45). This conjugation forms the physical basis for holding the four microspores together as one unit during the period of callose digestion. Sometimes, the distal pronexine appears flat and not entangled with the callosic envelope (Fig. 34). We believe this is an artifact because the hydrophilic nature of calloses makes preservation difficult and because no monad pollen is found in these two Annona species in this study or previously (Walker, 1971a ). A strong, specific conjugation between the pronexine of microspore and the callosic envelope appears in both species.

The third interesting feature is the microspore rotation, which results in the reversal of the proximal and distal walls of each microspore during the transition from tetrad to free spore stages. Such a reversal is made possible by the preexistence of the two phenomena mentioned above and by the further slow digestion of the remaining callose-cellulose structure; through the latter process the callose-cellulose structure is contracted and subsequently pulls the distal, thin pronexine of each microspore toward the center becoming the proximal side; consequently, the thick proexine becomes the distal side. The callosic envelope of the microspore tetrad is generally considered a transient structure; its role during the pollen development has been widely speculated but poorly understood (Buchen and Sievers, 1981 ; Raghavan, 1997 ; Majewska-Sawka and Rodriguez-Garcia, 1999 , for some recent reviews). In this study, a complicated, dynamic, and persistent callosic envelope is revealed in Annona glabra and A. montana. The functioning of this envelope as the binding system for the pollen tetrad and resulting in microspore rotation in Annona is interesting and unique, as far as we are aware, in the angiosperms.

Proexine formation, including how the apertures are determined and how those solid wall layers are built up, is a fascinating subject. Thirty years ago, Heslop-Harrison (1971) and Dover (1972) showed that the location of the pore is in close relation to the axes of the second meiotic division in some species. In this study the positioning of thick proexine and thin pronexine regions of the meiotic microspores is constantly correlated with the second meiotic division plane, which also strongly suggests that the proexine allocation in Annona glabra and A. montana is mainly determined during or right after the second meiotic division.

The details of proexine initiation and development vary considerably among different taxa, as that of exine development (Rowley, 1990 ). In Annona, the prosexine and pronexine are built up in their respective glycocalyx networks, and dictyosome vesicles are often found between the plasmalemma and the network. In most reports, including this one, the glycocalyx layer is developed synchronously with the proexine development both spatially and temporally, regardless of whether it forms a well-constructed network or loosely organized. The function of the glycocalyx layer in proexine formation has been emphasized (Rowley, 1975 ; Gabarayeva, 1993b ), especially chemically in its selective binding capacities. We believe its physical function to be very important as well, at least for Annona-like pollen with tall probacula and reticulate protectum. Building up solid proexine layers in a precise pattern in a liquid environment for a prolonged period of time surely requires certain physical support, and the glycocalyx layer is a crucial structure casting a stable environment, providing appropriate physical and chemical factors for sporopollenin polymerization.

With regard to the prosexine formation in Annona, SAPs are the structural units for both protectum and probacula. Those particles, about 40–50 nm in diameter by the time we could identify them, differ from the prosexine receptors in Asimina (Annonaceae; Gabarayeva, 1992 ), but resemble those of the prosexine in Anaxagorea (Annonaceae; Gabarayeva, 1995 ), Eucommia (Eucommiaceae; Rowley, Skvarla, and Pettitt, 1992 ), and Poinciana (Leguminosae; Skvarla and Rowley, 1987 ). In Annona, organization of SAPs during early probaculum development is highly specialized. Usually two SAPs stack into one radial row and five or six radial rows group in one whorl, which forms the forerunner of one probaculum (Figs. 14, 15). Then the SAPs of each group enlarge and fuse into one probaculum (Figs. 16–19). The stack of two SAPs in one radial file can also be found sporadically in Eucommia and Poinciana. In Annona the SAPs emerge and increase in number in the early phase of prosexine formation. Further development seems mainly to rely on the enlargement of those SAPs through active sporopollenin polymerization. Because the probacula are more or less uniform in their heights and diameters and no free SAPs or half probacula appear in the network later on, we consider that probacula initiation in Annona is restricted to the early phase of prosexine formation. After that, probacula will not be newly generated, i.e., the size but not the number of probacula increases during later prosexine development. Dr. J. R. Rowley of Stockholm University also considers that probacula are formed at one time only (personal communication).

The profoot layer in Annona is evidently initiated from the plasmalemma (Figs. 23–26); it forms at the interbacular regions first and then beneath the probacula; afterwards, this layer with probacula attached to its distal surface becomes well separated from the plasmalemma. The proendexine then extends from the profoot layer downward. The proendexine in the two Annona species is of the commonest structure in the plants, i.e., composed of white-line-centered branches or lamination (Rowley, 1995 ). In these two species, the proendexine is poorly developed at the proximal side and even much reduced or absent locally at the distal side (Figs. 32–34). The proendexine always appears segmented and wavy (Figs. 29, 33), which may be explained by Rowley's functional interpretations that the endexine may open up and close off in maintaining the supply routes (Rowley, 1995 ).

The Annonaceae have about 120 genera. Both the phylogeny and systematics of this family are highly controversial. Pollen morphology has been extensively surveyed in this family (Walker, 1971a, b ; Le Thomas, 1980, 1981 ; Le Thomas, Morawetz, and Waha, 1986 ; Waha and Morawetz, 1988 ), and palynology is considered significant in understanding the higher level relationships in Annonaceae by some authors (Doyle and Le Thomas, 1994, 1996, 1997 ). However, in comparison with the great diversities of the pollen features the information on pollen ontogeny is so scanty (Waha, 1987 ; Gabarayeva, 1992, 1993a, 1995 ) that interpretations on many features or the evolutionary trends of pollen characters are often hypothetical, or at least, lacking strong support. The determination of the aperture region has been a controversial subject in the Annonaceae. Cases of tetrad pollen, such as Annona, with reduced exine at the proximal region are termed catasulcate to cataulcerate (i.e., proximal aperture) by Walker (1971a, 1974) , as a secondary aperture by Morawetz and Hesse (1984), or as inaperturate in some other reports (Le Thomas, Morawetz, and Waha, 1986 ; Waha and Morawetz, 1988 ; Doyle and Le Thomas, 1994, 1996, 1997 ), in which the reduced proximal side is considered to have evolved as a secondary formation adapted to the tetrad development as a functional unit.

In the present study, the distal region of the meiotic microspores of these two Annona species is well termed the apertural region because, from the developmental viewpoint, the process of building up a thin pronexine at this region is comparable to the usual pattern of aperture formation and, from the functional viewpoint, the active cytoplasmic invaginations and evaginations at this side demonstrate its functional role as an aperture in nutrient transfer during development (Rowley and Skvarla, 1976 ). In addition, these reduced regions in mature tetrads do act as sites of pollen tube germination (Walker, 1971b ). Our current data indicate that these two Annona species still retain the ancestral nature of producing a distal aperture (anasulcate) and the pollen type of Annona is highly likely to have evolved from an anasulcate monad (Le Thomas, Morawetz, and Waha, 1986 ; Doyle and Le Thomas, 1996 ). In fact, anasulcate pollen is rather prevalent in the Annonaceae, and choosing anasulcate monad as the most primitive pollen type in the Annonaceae is well supported by morphological data (Le Thomas, Morawetz, and Waha, 1986 ; Doyle and Le Thomas, 1994 ) and molecular data (Doyle, Bygrave, and Le Thomas, 2000 ).

We hope the present study provides some new insights for interpreting the evolution of pollen aperture as well as the compound pollen in the Annonaceae. More ontogenetical studies are needed to fully understand the great diversity of pollen characters in the Annonaceae.

	FOOTNOTES

 
¹ The authors thank Prof. John R. Rowley for his encouragement, comments, and generosity in providing references; Ms. Chong-Ni Li from Chiayi Agriculture Experimental Station for helping with sample collection; and Dr. Wan-Nan Jane for technical consultation. This project is supported by Institute of Botany, Academia Sinica and National Science Council, Republic of China (NSC89-2311-B-001-039 and 89-2311-B-001-181).

² Author for correspondence (botsou{at}ccvax.sinica.edu.tw ; Fax: 886-2-27827954)

	LITERATURE CITED

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

 
Bhandari N. N. 1981 The microsporangium. In B. M. Johri [ed.], Embryology of angiosperms, 52–121. Springer-Verlag, Berlin, Germany

Blackmore S. P. Crane 1988 The systematics implications of pollen and spore ontogeny. In C. J. Humphries [ed.], Ontogeny and systematics, 83–116. Columbia University Press, New York, New York, USA

Buchen B. A. Sievers 1981 Sporogenesis and pollen grain formation. In O. Kiermayer [ed.], Cytomorphogenesis in plants, 349–376. Springer-Verlag, Berlin, Germany

Dahl A. O. J. R. Rowley 1991 Microspore development in Calluna (Ericaceae). Exine formation. Annales Sciences Naturelles, Botanique, Paris, 13^e serie 11: 155-176

Dover G. A. 1972 The organization and polarity of pollen mother cells of Triticum aestivum. Journal of Cell Science 11: 699-711[Abstract/Free Full Text]

Doyle J. A. P. Bygrave A. Le Thomas 2000 Implications of molecular data for pollen evolution in Annonaceae. In M. M. Harley, C. M. Morton, and S. Blackmore [eds.], Pollen and spores: morphology and biology, 259–284. Royal Botanic Gardens, Kew, UK

Doyle J. A. A. Le Thomas 1994 Cladistic analysis and pollen evolution in Annonaceae. Acta Botanica Gallica 141: 149-170[ISI]

Doyle J. A. A. Le Thomas 1996 Phylogenetic analysis and character evolution in Annonaceae. Bulletin du Museum national d'Histoire naturelle, section B, Adansonia 18: 279-334

Doyle J. A. A. Le Thomas 1997 Significance of palynology for phylogeny of Annonaceae: experiments with removal of pollen characters. Plant Systematics and Evolution 206: 133-159[CrossRef][ISI]

Dunbar A. 1973 Pollen development in the Eleocharis palustris group (Cyperaceae) 1. Ultrastrucutre and ontogeny. Botaniska Notiser 126: 197-254[ISI]

Fries R. E. 1959 Annonaceae. In H. Melchior [ed.], Die naturlichen Pflanzenfamilien, 2nd ed., 17aII, 1–170. Duncker & Humblot, Berlin, Germany

Gabarayeva N. I. 1992 Sporoderm developemnt in Asimina triloba (Annonaceae). I. The developmental events before callose dissolution. Grana 31: 213-222[ISI]

Gabarayeva N. I. 1993a Sporoderm developemnt in Asimina triloba (Annonaceae). II. The developmental events after callose dissolution. Grana 32: 210-220[ISI]

Gabarayeva N. I. 1993b Hypothetical ways of exine structure determination. Grana Supplement 2: 54-59

Gabarayeva N. I. 1995 Pollen wall and tapetum development in Anaxagorea brevipes (Annonaceae): sporoderm substructure, cytoskelon, sporopollenin precursor particles, and the endexine problem. Review of Palaeobotany and Palynology 85: 123-152

Gabarayeva N. I. J. R. Rowley J. J. Skvarla 1998 Exine development in Borago (Boraginaceae). 1. Microspore tetrad period. Taiwania 43: 203-214

Heslop-Harrison J. 1971 Wall pattern formation in angiosperm micro-sporogenesis. In Control mechanisms of growth and differentiation,. Symposium of the Society of Experimental Biology 25: 277-300. Cambridge University Press, London, UK

Kenrick J. R. B. Knox 1979 Pollen development and cytochemistry in some Australian species of Acacia. Australian Journal of Botany 27: 413-427[CrossRef]

Knox R. B. E. Friederich 1974 Tetrad pollen grain development and sterility in Leschenaultia formosa (Goodeniaceae). New Phytologist 73: 251-258

Knox R. B. C. A. McConchie 1986 Structure and function of compound pollen. In S. Blackmore and I. K. Ferguson [eds.], Pollen and spores, form and function, 264–282. Academic Press, London, UK

Le Thomas A. 1980 Ultrastructural characters of the pollen grains of African Annonaceae and their significance for the phylogeny of primitive angiosperms (first part). Pollen et Spores 22: 267-342

Le Thomas A. 1981 Ultrastructural characters of the pollen grains of African Annonaceae and their significance for the phylogeny of primitive angiosperms (second part). Pollen et Spores 23: 5-36

Le Thomas A. W. Morawetz M. Waha 1986 Pollen of palaeo- and neotropical Annonaceae: definition of the aperture by morphological and functional characters. In S. Blackmore and I. K. Ferguson [eds.], Pollen and spores, form and function, 375–388. Academic Press, London, UK

Majewska-Sawka A. M. Rodriguez-Garcia 1999 The special callose wall: a new insight. In C. Clement, E. Pacini, and J.-C. Audran [eds.], Anther and pollen, from biology to biotechnology, 119–128. Springer-Verlag, Berlin, Germany

Morawetz W. M. Hesse 1984 Primäre und sekundäre Aperturen bei Angiospermenpollen. In F. Ehrendorfer [ed.], Mitteilungsband der Botanikertagung in Wien. Kurzfassung der Beiträge

Raghavan V. 1997 Molecular embryology of flowering plants. Cambridge University Press, New York, New York, USA

Rowley J. R. 1975 The glycocalyx and initiation of exine spinules on microspores of Canna. American Journal of Botany 62: 479-485[CrossRef][ISI]

Rowley J. R. 1990 Are exine reception systems of Fagus and Pinus structurally the same? Diversity in exine development. Journal of Palynology 91: 323-344

Rowley J. R. 1995 Are the endexines of pteridophytes, gymnosperms and angiosperms structurally equivalent?. Review of Palaeobotany and Palynology 85: 13-34[CrossRef]

Rowley J. R. J. J. Skvarla 1976 Surface coating of germinal apertures of pollen and evolution of aperture. In G. W. Bailey [ed.], 34th Annual Proceedings of Electron Microscopy Society of America, 42. Electron Microscopy Society of America, Miami Beach, Florida, USA

Rowley J. R. J. J. Skvarla N. I. Gabarayeva 1999 Exine development in Borago (Boraginaceae). 2. Free microspore stages. Taiwania 44: 212-229

Rowley J. R. J. J. Skvarla J. M. Pettitt 1992 Pollen wall development in Eucommia ulmoides (Eucommmiaceae). Review of Palaeobotany and Palynology 70: 297-323[CrossRef]

Sampson F. B. 1977 Pollen tetrads of Hedycarya arborea J. R. et G. Forst. (Monimiaceae). Grana 16: 61-73

Sampson F. B. 1981 Synchronous versus asynchronous mitosis within permanent pollen tetrads of the Winteraceae. Grana 20: 19-23

Skvarla J. J. D. A. Larson 1963 Nature of cohesion within pollen tetrads of Typha latifolia. Science 140: 173-175[Abstract/Free Full Text]

Skvarla J. J. P. H. Raven J. Praglowski 1975 The evolution of pollen tetrads in Onagraceae. American Journal of Botany 62: 6-35[CrossRef][ISI]

Skvarla J. J. J. R. Rowley 1987 Ontogeny of pollen in Poinciana (Leguminosae) I. Development of exine template. Review of Palaebotany and Palynology 50: 293-311[CrossRef]

Takahashi H. K. Sohma 1980 Pollen development in Pyrola japonica Klenze. Scientific Report, Tohoku University, series IV (Biology) 38: 57-71

Takahashi H. K. Sohma 1982 Pollen morphology of the Droseraceae and its related taxa. Scientific Report, Tohoku University, series IV (Biology) 38: 81-156

Takahashi H. K. Sohma 1984 Development of pollen tetrad in Typha latifolia L. Pollen et Spores 26: 5-18

Waha M. 1987 Sporoderm development of pollen tetrads in Asimina triloba (Annonaceae). Pollen et Spores 29: 31-44

Waha M. W. Morawetz 1988 Pollen evolution and systematics in Annonaceae with special reference to the disulcate Australian endemic genera. Plant Systematics and Evolution 161: 1-12[CrossRef][ISI]

Walker J. W. 1971a Pollen morphology, phytogeography, and phylogeny of the Annonaceae. Contribution of the Gray Herbarium of Harvard University 202: 1-131

Walker J. W. 1971b Unique type of angiosperm pollen from the family Annonaceae. Science 172: 565-567[Abstract/Free Full Text]

Walker J. W. 1974 Aperture evolution in the pollen of primitive angiosperms. American Journal of Botany 61: 1112-1136

This Article Abstract Full Text (PDF) Submit a response Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Services Email this article to a friend Similar articles in this journal Similar articles in ISI Web of Science Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via ISI Web of Science (6) Citing Articles via Google Scholar Google Scholar Articles by Tsou, C.-H. Articles by Fu, Y.-L. Search for Related Content PubMed Articles by Tsou, C.-H. Articles by Fu, Y.-L. Agricola Articles by Tsou, C.-H. Articles by Fu, Y.-L.