Printer Friendly

Sporogenesis in bryophytes: patterns and diversity in meiosis.


Meiosis is an essential event in the life cycle of all sexually reproducing plants. It occurs in specialized cells (sporocytes) of the sporophyte (spore producing) generation and yields haploid cells that initiate the gametophyte (gamete-producing) generation. The four cells resulting from meiosis/cytokinesis become a tetrad of spores, each encased in a heavy sporopollenin wall, the sporoderm. In free-sporing plants (bryophytes, lycophytes, monilophytes), spores serve as dispersal units that can withstand long periods of drought before germinating into free-living photosynthetic plants each with a new assortment of genes. These independent haploid plants are gametophytes that produce gametes mitotically. Union of gametes (fertilization) brings haploid sets of chromosomes together and initiates the diploid sporophyte generation. This alternation of generations is a central theme in plant reproduction and evolution in which spores play an essential role. In heterosporous plants and seed plants, continued development of the gametophyte may take place within the spore wall resulting in an enclosed multicellular gametophyte, a process known as endospory.

In seed plants, microsporogenesis results in the pollen grain (male gametophyte) and megasporogenesis in the embryo sac (female gametophyte). In angiosperms, ultimate reduction of the gametophyte generation has taken place; the male pollen grain comprises only three cells and the female embryo sac commonly comprises seven cells or, in some cases, as few as four cells as in some early angiosperms (Williams & Friedman, 2004; Tobe et al., 2007; Rudall et al., 2008). The mature spores/microspores are among the simplest plants known yet with an enormously complex ultrastructure leading Heslop-Harrison (1972) to view pollen development as "morphogenesis in miniature".

In spite of great diversity, especially in the complexity of the sporoderm, the fundamental components of sporogenesis are the same in free-sporing plants, where the walled spores serve to disperse new plants, and in seed plants where the walled microspores have evolved into pollen grains for transmission of male gametes. Sporogenesis usually excludes discussion of megasporogenesis in the seed plants where resultant megaspores remain enclosed within surrounding tissues of the megasporangium (ovule) and, at least in angiosperms, lack an elaborate sporoderm. In all cases:

1. The sporocytes are isolated from vegetative cells. This is a variable process that usually, at minimum, involves production of expandable walls around each sporocyte. The sporocyte wall, usually consisting of callose or mucopolysaccharides, isolates cells programmed to enter the meiotic pathway from the influence of surrounding vegetative cells. This isolation apparently plays a key role is establishing conditions for sporogenesis (Heslop-Harrison, 1966).

2. The nuclear division is meiotic in which sets of chromosomes are reduced (generally from two sets to one) and results in four nuclei each containing a haploid set of chromosomes.

3. The sporocyte cytoplasm is quadripartitioned and distributed equally (with very few exceptions) to the spores/microspores of the tetrad. The regular partitioning of sporocytes into a tetrad of spores is one of the most distinct examples of precise geometrical division in the plant kingdom. This process appears endogenously controlled although there are extreme and poorly understood examples (e. g. grasses and sedges) in which external signaling almost certainly is involved.

4. A distinctly patterned multilayered spore wall containing sporopollenin develops around each spore/microspore. These walls are among the most elaborate and ornate found in the plant kingdom.

5. Maturation, often concurrent with finalization of the sporoderm, usually involves development of desiccation tolerance in preparation for spore dispersal in free-sporing plants. In pollen development, mitotic division of the haploid spore nucleus produces the reduced male gametophyte within the elaborate spore wall.

It is generally agreed that the early land plants arose from streptophycean algal ancestors (Graham, 1993; Becker & Marin, 2009; Finet et al., 2010) and that the early land plants, responsible for the initial greening of the earth and making the land habitable for life as we know it, resembled the bryophytes and that all higher land plants evolved from these pioneers. As remnants of the early land flora, bryophytes retain various cytological features that provide glimpses into the mechanisms utilized in the early achievement of sporogenesis. It is this evolutionary diversity that provides a foundation for determining fundamental developmental processes, their controls, and the possible applications in deciphering plant relationships. The underpinning of most, if not all, of the developmental processes responsible for plant growth and development are found in bryophytes. Comparative studies of sporogenesis including the mechanisms of meiosis, cytokinesis, and spore wall elaboration provide valuable information leading to a better understanding of evolutionary relationships among major groups of primitive land plants (Blackmore & Knox, 1990). We are now at a junction where it can be expected that information on sporogenesis will have an impact on our understanding of the control and process of plant development as well.

Plant microtubules underlie all phases of plant development, such as determination of division plane, cell shaping and wall deposition, in addition to mitosis/meiosis and cytokinesis. Unlike animal cells, where microtubules are nucleated at discrete centriole containing centrosomes, plant cells produced a bewildering assortment of microtubule arrays in the absence of centrosomes. The origin, development, and function of the diffuse anastral spindle of land plants are fundamental questions in evolutionary and developmental biology. Examples of discrete but centriole-less microtubule organizing centers (MTOCs) have been discovered in bryophytes (Brown & Lemmon, 1988a, 1990, 2007; Shimamura et al., 2004). The unparalled diversity of spindle organization and division patterns reeognized in bryophytes affords opportunity for investigations that will have far-reaching implications in understanding the nucleation and organization of plant microtubules and their role in morphogenesis.

One of the least understood aspects of sporogenesis is how the living cytoplasm of a single cell controls the intricate patterning of the sporopollenin-containing sporoderm. Sporopollenin is the most biologically resistant material on earth. It is produced by the cytoplasm at the cell surface within the sporocyte wall, but its synthesis and intricately patterned deposition remains little understood. Since the last comprehensive review of sporogenesis in bryophytes (Brown & Lemmon, 1988a), the evolution and development of spore and pollen walls has been the subject of recent reviews (Blackmore et al., 2007; Wallace et al., 2011).

On the other hand, considerable new information has accumulated on meiosis in bryophytes, particularly the highly diverse liverworts. These new discoveries are due largely to advances in techniques of immunofluoresence and confocal microscopy which have made possible the study of the cell biology of meiosis and simultaneous cytokinesis resulting in spore tetrads. Global views of the cytoskeleton during meiosis greatly enhance the interpretation of micrographs from transmission electron microscopy. We have accumulated data using both techniques (for specifics of methods, see Brown & Lemmon, 1995; Brown et al., 2007) and have attempted to use one to augment the other in order to achieve a better understanding of the process of sporogenesis in bryophytes. These studies of bryophytes have revealed a surprising diversity of meiotic mechanisms indicative of the adaptive radiation of the bryophyte-like early land plants.


The bryophytes (monosporangiates) are a natural and ancient group of plants that appear sister to protracheophytes and tracheophytes (polysporangiates). Bryophytes were among the original colonizers of terrestrial habitats and their status as the oldest living land plants is rarely contested (Mishler & Churchill, 1984; Shaw & Renzaglia, 2004; Renzaglia et al., 2007). Norton Miller has elegantly summarized the amalgamation of features that makes a plant a bryophyte as "a dominant and variously elaborated gametophyte; and a subordinate, non-append-aged sporophyte attached continuously throughout its life to the gametophyte and with a single terminal sporangium".

To say bryophytes are small is hot to imply they are simple for they are hot. Although individual plants (ramets) may be small, they frequently produce massive colonies; their biomass reaching an extreme in the peat forming sphagnum mosses (Miller, 1993). Bryophytes are an extraordinarily diverse and successful group. They are non-vascularized and have a life cycle dependent upon close association with water. Although they reach the greatest species diversity in tropical and temperate rainforests, they are cosmopolitan and are round in every terrestrial habitat; epilithic, epiphytic, terrestrial and aquatic. They were early pioneers of life on land and today are still adapted to life in habitats without soil and are often weedy, revegetating sites of natural, or more often man made, disturbances. Vegetatively bryophytes show remarkable desiccation tolerance not found in other plant groups in order to weather the ever-present threat of drought.

The outstanding character of one sporophyte/one sporangium per fertilization event has lead to thumbnailing these plants as "monosporangiates". They differ from all other plants ("polysporangiates") in the determinate nature of the sporophyte, which is attached to and partially dependent upon the gametophyte throughout its life history. Developmentally, it follows that the sporophyte meristem is quite different in being both determinate and non-branching. Whether there are simple, short-lived meristems at the apices is debatable. This has lead some authorities to question if the meristem is the functional equivalent of that in vascular plants or that it is the division and differentiation of a subapical group of derivatives that produces the entire sporophyte (Crandall-Stotler, 1980). It will be interesting to compare the pattern of gene expression in "meristems" of bryophytes and tracheophytes.

The sporophyte consists of a basal foot, setum or connective stalk supporting a single sporangium or capsule. A conspicuous placenta develops at the gametophyte/ sporophyte junction. The lower portion of the sporophyte is referred to as the foot and is deeply inserted into the surrounding gametophyte tissue, which forms a multi-layered sheath or vaginula around it (Ligrone & Gambardella, 1988). The interacting cells develop a complex wall labyrinthine characteristic of transfer cells (Pate & Gunning, 1972). Studies of the moss placenta were significant in physiological and autoradiographic experiments demonstrating the function of transfer cells in facilitating transport of photosynthate in plants (Browning & Gunning, 1979a, b). The placenta (modified gametophyte-sporophyte junction) is an autapomorphy of land plants (Graham, 1996; Graham et al., 2000) and its variations have proven important in phylogenetic studies (Carafa et al., 2003).

Although the sporophyte is small and dependent, elaborate structures have evolved to optimize spore dispersal. The moss peristome is a very complicated structure with no parallel in the plant kingdom. The peristome, which surrounds the mouth of the sporangium, consists of hygroscopic teeth that regulate the dispersal of mature spores according to optimal moisture conditions. The other groups of bryophytes have totally different mechanisms for aiding spore dispersal. Liverworts produce unique cells, the elaters, which are elongated cells with spiral thickenings that are hygroscopic and contract/expand with moisture changes. Hornworts have no specialized structures to aid spore dispersal but a unique intercalary meristem results in relatively long period of spore production. A continuous supply of mature spores are elevated by growth of the sporangium and released by terminal slits in the sporangium wall.

It is generally thought that there are about 20,000 species of bryophytes in three major groups: 10-14,000 mosses, 5,000 liverworts and 400 hornworts. Although diverse, ancient, and ecologically important, the plants are of little direct importance to man and easily overlooked, perhaps in part accounting for the lack of biological information about this important group of organisms. Much remains to be discovered about the cell and molecular biology of bryophytes, especially of the biochemistry of spore walls and the diversity of oils and carbohydrates, as well as the regulation of developmental processes. Although there is a tendency to recognize the three major groups of bryophytes as separate Divisions (e. g. Goffinet & Shaw, 2009), we believe the classification of Crum and Anderson (1981) of the bryophytes as a single Division with three subdivisions--liverworts (Marchantiophytina), mosses (Bryophytina), and hornworts (Anthocerophytina)--best reflects the overall relationships of monosporangiates.

Perhaps nowhere has the rigorous application of molecular data in phylogenetic studies had more impact than in bryophyte classification (Frey & Stech, 2005; Forrest et al., 2006; Renzaglia et al., 2007; Crandall Stotler et al., 2009a, b). Although phylogenetic studies have supported monophyly of the three groups, there is no agreement as to which is the earliest divergent group. Because of several characters, especially the lack of stomates and the presence of the growth regulator lunularic acid, the liverworts are often regarded as sister to all other land plants. Because hornworts have wall xylans, stomates, and a well-differentiated sporophyte meristem similar to those of tracheophytes, there is reason to suspect they are closer to the tracheophytes. Major discoveries are still being made, e.g. Takakia, one of the more primitive of all mosses was first discovered in 1951 (Hattori & Mizutani, 1959) but details about the sexual phases of its life cycle were not known for another 40 years (Smith, 1990).

Realignments of relationships have occurred in each of the major clades. Perhaps of most basic importance is the recovery of early divergent forms representing remnants of early evolutionary adaptive radiation. As the result of molecular analysis, the Treubiales and Haplomitriales (Haplomitriopsida) are now recognized as sister to all other liverworts and the Blasiales are sister to the Marchantiidae in a monophyletic Marchantiopsida. In the mosses, the several non-peristomate groups (Andreaea, Sphagnum and Takakia) are confirmed as early divergent forms and Oedopodium is now recognized as sister to the predominate clade of peristomate mosses. Relevant here is the prevalence of trilete spores among these early mosses (Brown et al., 1982a, b; Renzaglia et al., 1997; Shimamura & Deguchi, 2008).

The phylogenetic relationships within the subdivisions as currently understood are detailed in the recent comprehensive reference book on the biology of bryophytes (Goffinet & Shaw, 2009). Although there will no doubt be further changes, especially as data from comparative studies of cell and developmental biology and morphology are added, the current phylogeny is valuable as a framework for discussing the unexpectedly complex and diverse process of sporogenesis in bryophytes.

Spores and the Early Evolution of Land Plants

The origin of sporogenesis is central to the evolution of land plants from putative algal ancestors. Whereas the meiospores (the haploid cells resulting from meiosis) of land plants are covered with desiccation resistant sporopollenin walls, the meiospores of algae are not. It is the zygote of algae that becomes covered with a resistant wall. All land plants (embryophytes), including the bryophytes, have a well-defined alternation of two generations with meiosis occurring within distinctive sporangia of the sporophyte. It was generally assumed that the alga-plant transition required the evolution of a sporophyte before spores were produced (e.g., Graham, 1993; Kendrick and Crane, 1997; Becker & Marin, 2009). However, this scenario is not supported by fossil evidence and is further called into question by the cell biology of meiosis/spore cleavage in extant bryophytes. We have developed a concept of the origin of sporogenesis independent of the origin of alternation of generations (Brown & Lemmon, 2011a).

Analysis of data on meiosis in all groups of bryophytes has led us to hypothesize that the sporopollenin wall typically covering the algal zygote was transferred directly to meiospores without an intervening sporophyte generation. The heterochronal events involved would be an acceleration of meiosis and a concurrent delay in wall deposition until after meiosis produced the tetrad of meiospores. In this scenario, the land could have been populated by extremely simple thalloid or even filamentous gametophytes that produced gametes and zygotes. Each zygote immediately underwent meiosis resulting in a tetrad of spores. Obviously the eventual intercalation of mitosis before meiosis would increase the number of cells entering the meiotic pathway, but the critical transition from aquatic to terrestrial habitat could have been made possible by the expedient transfer of resistant wall from zygote directly to meiospores. On land, the production of four spores from each zygote, capable of both survival and dispersal, would be immeasurably more valuable than any number of meiospores designed for dispersal in water and doomed to perish on land. We can assume that the reproductive advantage of many fertilization events was thus realized by dissemination of walled spores ready to germinate into new plants on land. The dispersal of spores increases the potential for sexual recombination by establishing new populations of gametophytes in close proximity to one another, thus increasing the likelihood of gene flow. Even before vegetative bodies developed strategies to resist drying, the earliest progenitors of land plants could have survived because of the sexual reproductive advantages of spores, serving as both survival and dispersal units. We present two lines of evidence, the fossil record and cell biology of sporogenesis in bryophytes, upon which we base this interpretation.

The Fossil Record

The first evidence of life on land is a sudden abundance of sporopollenin walled spores without any indication of the plants that produced them. This early period, recognized by Gray (1993) as the Eoembryophytic, was a long one lasting from approximately 476 to 432 million years ago. It is generally agreed (e.g. Graham, 1993, 1996; Finet et. al, 2010) that the early land plants, which were responsible for the initial greening of the earth and making the land habitable for life, resembled the bryophytes and that all higher land plants evolved from these pioneers. It is also generally agreed that the early land plants arose from charophycean algal ancestors that typically have only one diploid cell in the life cycle, the zygote. The zygotes become covered with a heavy sporopollenin wall and serve as over-wintering survival units.

The fossil record has been valuable not only in convincing us that spores were produced by the pioneering land plants, but also that sporogenesis continued to evolve over an initial period of 40-50 million years. Blackmore and Barnes (1987) hypothesized that two factors contributed to evolution of the embryophyte spore wall; transfer of the sporopollenin wall from zygotes to meiospores, and the fusion of single lamina into multilaminate walls. Putative intermediates in the evolution of the multilaminate spore wall have been identified in Cambrian assemblages (Taylor & Strother, 2008).

The oldest palynomorphs, designated cryptospores, constitute an assemblage of monads, dyads, and tetrads, all of which remain attached and enclosed in a tight-fitting envelope (synoecosporal wall). After this initial period, monads, dyads and tetrads without envelopes appeared and were followed by free hilate monads and trilete spores.

No fossil evidence of the plants/plant progenitors that produced this diversity of spores has been found. Although it is generally agreed that the early plants were at the bryophyte grade of organization, there is no fossil evidence of their morphology. The first plant megafossils occur 65 million years after the earliest spores appeared and they are usually interpreted as primitive vascular plants (protracheophytes). The first bryophyte fossils are liverwort-like from the Devonian and contemporaneous with tracheophytes (Edwards et al., 1995). Sporogenesis in extant bryophytes retains ontogenetic clues that help make sense of the great enigma of a fossil record of millions of years of spores without fossils of the protoembryophytes that could have produced them.

Cell Biology of Sporogenesis in Extant Bryophytes

Plants of the bryophyte grade are living treasures that contribute to the understanding of how sporogenesis might have evolved at the algal-plant transition. Meiosis in bryophytes is characterized by a seemingly curious reversal of the usual sequence of nuclear division followed by cytokinesis. The sporocytes of most bryophytes become quadrilobed leaving the undivided diploid nucleus in the central cytoplasm (Brown & Lemmon, 1988a). We interpret this precocious initiation of cytokinesis before meiosis as a clue to the mechanism for transfer of wall deposition from zygote to meiospores. The precocious initiation of cytokinesis, if occurring in algal-plant transitional forms, could in itself account for cryptospores, the oldest fossil spores, which seem to be imperfectly divided tetrads enclosed by a synoecosporal wall.

Sporogenesis in bryophytes indicates that transfer of the sporopollenin wall from zygote to meiospore was accomplished by both precocious quadrilobing of the cytoplasm into the four future spore domains and a delay in timing of the deposition of the complex spore wall (sporoderm). There is evidence among living bryophytes that wall deposition begins in meiotic prophase in some liverworts, and tetrads remain intact in others. Certain liverworts have wall precursors covering the lobes of sporocytes in meiotic prophase (Brown et al., 1986; Brown & Lemmon, 2011a). In these examples, the sporocytes have four well-defined spore domains covered by prepatterned exine precursors before chromosomal division, i.e. on the diploid cell. Other liverworts (e.g. some species of Sphaerocarpos and Riccia) produce a permanently adhered tetrad of spores. The ornamentation develops over the entire circumference of the spore tetrad and is continuous from one spore surface to the next. Unlike free tetrads of other bryophytes, germination is from the unspecialized distal surfaces (Proskauer, 1954). In mosses, hornworts, and some liverworts, precocious quadrilobing is associated with monoplastidic meiosis. In this distinctive type of meiosis, quadripolarity is established in early prophase by two divisions of a single plastid. Migration of the four plastids to tetrahedral positions define spore domains, regions of cytoplasm that will each receive a nucleus at the completion of meiosis. The four plastids serve as plastid-MTOCs and cones of microtubules emanating from them interact to further define cytokinetic planes and initiate a quadripolar spindle.

As we now know, this very striking example is only one manifestation of the coordination of microtubule organization with the control of division plane in bryophyte meiosis. Quadilobing in certain liverworts is preceded by pre-meiotic bands of microtubules marking the future cytokinetic planes before the nucleus enters prophase I (Fig. 1). These processes contribute to wall deposition on spores, or at least spore domains, as opposed to surrounding the entire zygote as in algae.

Evidence suggests that once cytokinesis was initiated prior to meiosis, a quadripolar spindle was required for coordination with the predetermined spore domains. Diversity of meiotic spindle ontogeny is a characteristic of early land plants that evolved in association with the origin of sporogenesis. The diversity in spindle origin seen in bryophytes is in stark contrast to the uniformity of anastral spindle origin in "higher plants" and centrosomal spindle origin in algal and animal cells.

Bryophytes thus approach the process of sporogenesis as a single comprehensive event in which quadripolarity is strongly expressed in the early sporocytes. This is evidenced by: 1) diploid cytoplasm becomes quadrilobed, sometimes so deeply that only the central portion containing the nucleus remains undivided, 2) the first division spindle is initially quadripolar, and 3) wall precursors may be present at the periphery of the cytoplasm as early as prophase of first meiosis.

Microtubules and their Organization in Plants

All aspects of plant development are dependent upon the cytoskeleton responding to cascades of developmental signals. The plant cytoskeleton consists principally of microtubules and microfilaments, molecules that are remarkably conserved in eukaryotic organisms. The most conspicuous elements in plant cells are microtubules. These helical filaments composed of alpha and beta subunits of tubulin tend to be in a constant state of flux, being assembled, degraded, and re-assembled into different arrays. Although the cells of plants and animals share many basic features, the encasement of plant cells within a rigid cell wall has led to unique microtubule systems that have no counterpart in animal cells (Brown & Lemmon, 2001; Wasteneys, 2002).

Two orderly cycles of microtubules occur during the plant life cycle; one in vegetative growth where new cells become part of a tissue, and the other in reproduction where cells function as individual entities (Brown & Lemmon, 2001). In vegetative cells, hoop-like cortical microtubules underlie the cell wall during interphase. As cells prepare to enter mitosis cortical microtubules are replaced by the pre-prophase band (PPB). The PPB is component of the cytokinetic apparatus (Gunning, 1982) that marks and prepares the future division site. The PPB disappears as the mitotic spindle is organized. After chromosomes move to poles in anaphase, the phragmoplast mediates cell plate formation that completes cytokinesis. In cells of the reproductive cycle, a nuclear-based radial microtubule system (RMS) replaces the hoop-like cortical system and PPBs are absent. It is probable that no land plant completes its life cycle without a switch at one point or another to this microtubule cycle. The two plant microtubule cycles relate directly to the determination of wall placement and are referred to as the PPB cycle and the RMS cycle.

The organization of highly polar microtubules into functional arrays occurs at sites known as microtubule organizing centers (MTOCs). First postulated as essential entities in plant cells lacking centrosomes (Pickett-Heaps, 1969), the illusive plant MTOC is now known to be composed of [gamma]-tubulin and [gamma]-tubulin associated proteins. [gamma]-Tubulin is universally associated with nucleation of microtubules in eukaryotic cells (Ovenchkina & Oakley, 2001; Schmit, 2002; Binarova et. al., 2006; Wiese & Zheng, 2006). [gamma]-Tubulin in land plants has been released from a tight association with centrioles that is typical of animal and algal cells. Typically, plant [gamma]-Tubulin is a pleiomorphic entity that moves in a cell-cycle specific manner to different locations where it nucleates the microtubule systems that drive cell division and morphogenesis (Brown & Lemmon, 2007). The double staining of cells with indirect immunofluorescence to label [gamma]-Tubulin and microtubules in separate channels has shown that [gamma]-Tubulin is routinely present at the minus-ends of microtubules where assembly occurs and thereby serves both as a marker of the MTOC and polarity of the microtubules.

Mitotic spindle initiation in the three major taxa of bryophytes is an important aspect of the evolution of land plants from putative algal ancestors. This information is important in understanding the divergence of plant lineages, and in tracing the evolution of the plant MTOC and its role in organizing the anastral plant spindle. This is not an issue in animal evolution; centriolar centrosome based spindles are present uniformly throughout the animal kingdom, except in rare cases (usually in egg development) where spindles may form in the absence of centrosomes. In plants, the absence of centrosomes is ubiquitous, except where they appear in spermatogenesis of plants such as the bryophytes that produce motile sperm (Shimamura et al., 2004). This exquisite developmental control over the expression of centriolar centrosomes in the life cycle of these plants could either have been inherited from algal ancestors or have been acquired as an adaptation to life on land.

Spindles are initiated at different types of MTOCs in the three major lineages of bryophytes (Brown & Lemmon, 2011b) (Fig. 2). The mitotic spindle of liverworts is organized by unique polar organizers (POs) (Fig. 2a-c). Although superficially similar to the centrosomes of algal and animal cells, POs are not permanent cell organelles and they do not contain centrioles. Rather, they are transient concentrations of [gamma]-Tubulin that appear de novo at opposite poles of nuclei preparing to divide and disappear by metaphase (Brown et al., 2004). Mitotic spindles in mosses are anastral (Schepf, 1984; Doonan et al., 1987; Brown & Lemmon, 2011b). They are organized by [gamma]-Tubulin associated with the nuclear envelope (NE-MTOC) (Fig. 2d-f). Spindle origin in the hornworts (Fig. 2g-i) is associated with division of the single plastid, which serves as an MTOC (P-MTOC). A unique axial microtubule system (AMS) that forms in association with division of the plastid gives rise to the mitotic spindle (Brown & Lemmon, 1988b, 2011b; see also the section on hornworts). Regardless of origin, mature metaphase spindles in all three groups of bryophytes are indistinguishable from the typical anastral spindle of higher plants (Fig. 3). [gamma]-Tubulin spreads into a rounded cup at the poles and the spindle consist of numerous subsets of converging microtubules called minipoles by Schepf (1984). A curious phenomenon of plant spindles, true of bryophytes and seed plants, is the presence of [gamma]-Tubulin in the spindle fibers themselves (Liu et al., 1993; Schmit, 2002; Shimamura et al., 2004; Brown & Lemmon, 2007) (Fig. 3b, e, h).

After chromosomes move to opposite poles, [gamma]-Tubulin migrates from polar/distal surfaces of reforming nuclei to the proximal surfaces where it generates the opposing arrays of phragmoplast microtubules (Fig. 4). This repositioning of MTOCs that generate opposing microtubules in the interzone also occurs after mitosis in higher plants (Brown & Lemmon, 2007).

All three types of MTOCs (POs, P-MTOCs, and NE-MTOCs) have been discovered in meiosis of bryophytes (Brown & Lemmon, 2007; Brown et al., 2010). All of these MTOCs are known to organize the initially quadripolar meiotic spindles in bryophyte sporogenesis. In mosses and homworts, quadripolar spindles are organized exclusively by P-MTOCs. In liverworts, all forms of MTOCs occur (Brown & Lemmon, 2006, 2009; Brown et al., 2010) suggesting an adaptive radiation involved with the innovation of sporogenesis in early land plants.

At the transition from archesporial cells to sporocytes in mosses and hornworts, special expandable walls of callose or mucopolysaccharide are formed within the vegetative cell walls. The surrounding cellulosic cell walls lyse and the released sporocytes depart on a pathway driven by the RMS system lacking cortical and PPB microtubules. With rare exception, the sporocytes of hornworts, mosses and liverworts establish quadripolarity in anticipation of division into four spores prior to meiosis; the future spore domains of the tetrad are defined early in prophase I. It only remains for the spindles of the first and second meiotic divisions to deliver nuclei into the four predetermined spore domains. Typically, the meiotic spindle is quadripolar in origin. It becomes functionally bipolar by convergence of poles in pairs toward a single spindle axis that terminates at cytokinetic furrows. The second division spindles deliver nuclei into the four spore domains.

Just as the PPB in mitosis shows that the cytoplasm is divided into two domains before nuclear division, the establishment of quadripolarity in sporocytes shows that the four future spore domains of the tetrad are determined before meiosis. Whereas we usually think of a cytoplasmic domain being organized around a nucleus, a nuclear cytoplasmic domain or NCD (Brown & Lemmon, 1992), in bryophyte meiosis the domain is one that will receive a nucleus after nuclear division. This is similar to the situation in animal cells where the centrosome divides and moves to opposite poles and two cytoplasmic domains are established before nuclear division (Mazia, 1961). Thus, the cytoplasmic domain is a unifying concept in cell organization and division in both plants and animals.

The most outstanding examples are found in sporogenesis of plants with monoplastidic meiosis and simultaneous cytokinesis. The behavior of plastids and the quadrilobing of cytoplasm alerted early light microscopists to the unusual manifestation of predetermined wall placement in monoplastidic sporocytes. Bradley Moore Davis (1899) pondered, "Can it be supposed that cytoplasm would be entrusted with so important a task as the preparation of a chloroplast for each of the four nuclei that are later to preside over the spores before there is any indication that such nuclear division is to take place?" It was the advent of microtubule visualization by TEM and indirect immunofluorescence, however, that allowed investigators to partially answer this question and fully appreciate the complexity of the unique multipurpose cytoskeleton intimately involved in the spatial and temporal coordination of meiosis and cytokinesis in bryophyte sporogenesis (Brown & Lemmon, 1982a, b, 1987a, b; Busby & Gunning, 1988a, b, 1989; Shimamura et al., 2004).

Quadripolarity and the Quadripolar Mierotubule System

Cells entering the meiotic pathway no longer have hoop-like cortical microtubules and typically lack PPBs. The sporocytes of bryophytes, however, clearly define the future spore domains before meiosis distributes four haploid nuclei into the tetrad members. The sporocytes become quadripolar as they enter the meiotic pathway. In some, the cytoplasm becomes deeply lobed and in others the cytoplasm remains spherical with only shallow furrows marking the four future spore domains. This early establishment of quadripolarity clearly indicates that meiosis is approached as a single event rather than two sequential divisions. The only examples of "apolar" sporocytes, i.e. without prophasic quadripolarity are found in the complex march-antioid liverworts where meiosis often results in variable tetrad arrangements (Brown & Lemmon, 1988c; Shimamura et. al., 2011). In apolar sporocytes of Conocephalum conicum, the spindles are variously oriented and spore domains are determined after meiosis via RMSs emanating from the tetrad nuclei (Brown & Lemmon, 1988c). Post-meiotic determination of division planes is more typical of sporogenesis in tracheophytes, except for some lycophytes and eusporangiate ferns that undergo monoplastidic meiosis (Brown & Lemmon, 1997).

It has recently been discovered (Brown & Lemmon, 2006, 2009) that quadilobing in certain liverworts is preceded by pre-meiotic bands of microtubules marking the future cytokinetic planes before the nucleus enters prophase I. All Jungermanniopsida so far sampled exhibit the pre-meiotie bands (Brown & Lemmon, 2006, 2009) as does the basal marchantioid Blasia (Brown et al., 2010). Although the precise function of the bands is not well understood, it is clear that they play a role in quadrilobing. They may control quadrilobing by restricting cytoplasmic growth thereby causing lobing between cytoplasmic furrows and/or modifying the underlying membrane in the furrows somehow preparing the division sites or controlling vesicle trafficking to the sites. These factors would establish quadripolarity and promote sporoderm wall deposition on spores, or at least spore domains, as opposed to surrounding the entire diploid cell. We further suggest that the pre-meiotic system of microtubule bands involved in precocious quadrilobing of the sporocyte cytoplasm may have evolved in early land plants and was (Brown & Lemmon, 2011a) transmitted to the later evolving sporophyte generation as a mechanism for control of division polarity in vegetative growth. The early establishment of division plane involving girdling bands of microtubules perhaps has its roots in the streptophycean algae, e. g. Stichococcus and Klebsormidium (Pickett-Heaps, 1974). The mitotic PPB is today a distinguishing feature of land plants that marks the future cytokinetic plane before prophase (Gunning, 1982; Mineyuki, 1999, 2007; Brown & Lemmon, 1993, 2001, 2007; Wasteneys, 2002). It may be that the PPB is a superior developmental program that confers precise control of wall placement and thereby contributed to elaboration and the eventual dominance of sporophytes on land.

In mosses, hornworts, and some liverworts, precocious quadrilobing is associated with monoplastidic meiosis, a pleisiomorphy of land plants (Graham, 1993). Monoplastidic meiosis is found in all major clades of plants with the exception of spermatophytes (reviewed by Brown & Lemmon, 1997). In this distinctive type of meiosis, quadripolarity is established in early prophase by two divisions of a single plastid and migration of the four plastids to tetrahedral positions in the future spore domains. The four plastids serve as MTOCs and cones of microtubules emanating from them interact to form a quadripolar microtubule system (QMS) (Fig. 5a) further defining cytokinetic planes before eventually contributing to the development of the quadripolar meiotic spindle. Although it was previously thought that the QMS was exclusively associated with monoplastidic meiosis (Brown & Lemmon, 1993), we now know that this very striking example is only one manifestation of the coordination of microtubule organization with the control of division plane in meiosis. As can be seen in the many liverworts, the organization of astral systems from four distinct polar organizers also leads to a spectacular QMS (Fig. 5b).

All forms of bryophyte MTOCs are known to organize the quadripolar spindles in sporogenesis. In mosses and hornworts, quadripolar spindles are organized by P-MTOCs (Fig. 5a). In liverworts, often considered to be the earliest divergent land plants, all forms of MTOCs (POs, P-MTOCs, and NE-MTOCs) occur (Brown & Lemmon, 2006, 2008, 2009; Brown et al., 2010) suggesting again that many attempts were made to solve the problems involved with the innovation of sporogenesis. The diversity of meiotic spindle ontogeny is a characteristic of early land plants that probably evolved is association with the origin of sporogenesis.

Although unorthodox, the quadripolar spindle functions to distribute chromosomes normally. The four poles of the spindle serve to properly orient spindle axes with respect to the predetermined spore domains. The quadripolar spindle forms from four cones of microtubules emanating from MTOCs located in the four spore domains (Fig. 5). The plus end interaction of these opposing systems forms a QMS that encloses the nucleus in prophase I. The nuclear envelope breaks down in late prophase and as homologues attach to microtubules of the QMS, the poles converge in pairs toward a single spindle axis that terminates at opposite cleavage furrows. Thus, the mature metaphase I spindle is functionally bipolar with a pair of poles straddling opposite cleavage furrows. If the QMS arose from P-MTOCs, the spindle poles may slide away from the fixed plastids, the plastids themselves may move, or in one case (Sphagnum) the P-MTOCs remain intact and little or no convergence of poles occurs (Brown & Lemmon, 1993; see section on Sphagnum). In this case, the two polar-regions extend between the plastids of each pair.

Quadripolar spindles regularly result in telophase I nuclei located at, often curving around, opposite cleavage furrows. Rearrangements in the cytoskeleton at this stage continue to show the influence of quadripolarity. Microtubules emanate from the four poles (plastids or POs) generating four radial arrays centered in the four spore domains. Where plus ends of the microtubules interact they form stabilized arrays which are bi-polar in nature (Fig. 6). Subsets of this array appear spindle-like. This brings up the important point that a fundamental organization of plant microtubules, revealed by comparative studies of bryophyte sporogenesis, is a bipolar microtubule array (BMA). The best-known examples are the spindle and phragmoplast, but similar constructions can be seen in other phases such as the early bipolar axial microtubule systems associated with the dividing plastids (Fig. 6a, b), in the QMS itself (Fig. 6c, d), and in multipolar assembly of the phragmoplast in monoplastidic sporocytes (Fig. 6e, f).

The fundamental nature of the BMA was described (Brown & Lemmon, 1988b, 1993) in the unique AMS of homwort mitosis where the AMS orients parallel to the long axis of the dividing plastid and spindle and at right angles to the division plane and cell plate. This basic structure can explain the construction of the far more complex QMS in meiosis. In both hornworts and mosses, AMS-like arrays accompany the two rounds of plastid division (Fig. 6a, b). When the resulting four plastids reach the poles, they serve as MTOCs for the QMS which consists of distinct spindle-like BMAs interconnecting the four plastids in tetrahedral arrangement (Fig. 6c, d), the site of their plus end interaction coinciding with and defining the division cleavage planes and boundaries of domains.

Similar BMAs contribute to the development of complex phragmoplast systems (Fig. 6e, f). Although phragmoplasts typically form between the nuclei after first division, they disperse without directing wall deposition and second division occurs in the undivided sporocyte. The two spindles of meiosis II develop by interaction of microtubules from poles in the spore domains, the prophase spindles themselves are striking examples of BMA, (for example, see Fig. 45i). Thus, the quadripolar sporocyte achieves proper placement of tetrad nuclei, one in each of the predetermined spore domains. Phragmoplasts form among the four haploid nuclei, first between the daughter nuclei (primary) and then between non-daughters (secondary). As phragmoplasts expand they forma complex that guides deposition of walls and the tetrad of spores is simultaneously cleaved along the predetermined cytokinetic planes.

Spore Cleavage

Spore cleavage is a terre commonly used for cytokinesis following meiosis when the cytoplasm of the sporocyte divides into a tetrad of spores. There are several terres associated with the process of meiotic cytokinesis that are entrenched in the literature and thus have historical validity but can be confusing as they are used for both plant and animal cells. Terres such as cleavage furrows (Farr, 1916), constriction furrows (Fart, 1916), centripetal furrows (Furness & Rudall, 1999), and cleavage in furrowing (Furness et al., 2002) are apt to be equated with cytokinesis in animal cells. There is no evidence whatsoever that any of the variations of cytokinesis now recognized in plant cells (Otegui & Staehelin, 2000) is accomplished by actomyosin-driven constriction typical of cytokinesis in animal cells. Instead, meiotic cytokinesis, like plant cytokinesis in general, is completed by a phragmoplast-mediated secretory process of cell plate deposition. Meiotic cytokinesis in bryophytes typically follows precocious infurrowing of the cytoplasm, which is itself essentially a plant-like process involving differential growth of the cytoplasm and/or increased deposition of sporocyte wall material in the cleavage planes.

As has been previously stated, meiotic cytokinesis in bryophytes is almost always simultaneous and reflects the division planes established by the phenomenon of sporocyte quadrilobing. An exception is Amblystegium riparium, a moss that clearly undergoes successive cytokinesis (Brown & Lemmon, 1982b) and there are vague references to successive division in liverworts. However in these cases the sporocyte is so deeply lobed that it is very difficult to be certain that a dyad wall actually forms and to some degree it is irrelevant as the division sites are clearly determined and the sporocyte nearly cleaved in prophase. It is quite likely that successive division occurs in the closely related liverworts Aneura and Riccardia. However, the isthmus is so narrow that mechanical fracture can break the sporocyte into dyads as seen in preparations of Aneura pinguis (e.g. Fig. 52g, h, k, 1) where the enclosing sporocyte wall has been removed enzymatically. It should be noted that no examples of a dyad wall have been confirmed with TEM and what TEM is available has shown the formation of an organelle band marking the boundaries of the dyad domains rather than intersporal septa. Such artifactual dyads may be relevant in interpreting the fossil palynomorphs.

Even though meiotic cytokinesis is delayed until after second division when it occurs simultaneously, a well-developed phragmoplast is typically organized in the equatorial region during anaphase/telophase I. [gamma]-Tubulin becomes concentrated at the tightly focused polar regions in anaphase-telophase, then moves from distal (polar) to proximal surfaces of the sister groups of chromosomes where opposing arrays of microtubules are organized in the interzones between sister nuclei. The phragmoplast expands centrifugally to join with the pre-established division furrows. Instead of the phragmoplast directing deposition of an intersporal septum separating the dyad domains, organelles invade the equatorial region where they comprise a conspicuous sheet know as the organelle band (OB).

OBs have been reported after first meiosis in all major clades of land plants from bryophytes to orchids (Rodkiewicz et al., 1986; Brown & Lemmon, 1988a, 1991, 2001; Furness et al., 2002). The OB seems to be a meiotic component involved in inheritance of the organelles as part of the general program of definition of spore domains. The OB consists principally of mitochondria, liposomes, smaller organelles and various elements of the endomembrane system. In polyplastidic sporocytes, plastids are present in the organelle band whereas in monoplastidic sporocytes, the plastids remain at the tetrad poles where each will be inherited by a spore following deposition of intersporal plates after meiosis.

The mechanism of intracellular motility involved in OB development is not at all understood, especially since the organelles first appear to form a collar surrounding the interzone and then proceed to invade the region at right angles to the BMAs making up the phragmoplast. An interesting variation occurs in Equisetum where the plastids and mitochondria form distinct layers in the OB (Bednara & Rodkiewicz, 1984; Bednara et al., 1986). Following meiosis II in Equisetum, the organelles repopulate the spores in successive waves, first plastids, followed by mitochondria resulting in approximately equipartitioning of the organelles to each spore of the tetrad. It seems probable that this movement is dependent upon the microtubule systems that radiate from the tetrad nuclei.

Phragmoplast formation after second meiosis is complex and is mediated by a quadripolar structure of merged phragmoplasts similar in appearance to the initial QMS. The sequence of events is as follows: primary phragmoplasts develop in the interzones between pairs of telophase II nuclei, secondary phragmoplasts develop adventitiously among all non-sister nuclei and all six phragmoplasts merge. Cell plates are initiated simultaneously in the mid-zones of all six phragmoplasts. The cell plates in sporocytes tend to be deposited all along the raid-zones of phragmoplasts that occupy the cytoplasm between furrows. Spores in tetrahedral arrangement are cleaved in this manner and each spore has a tri-planar proximal surface. An exception to this general plan in bryophytes occurs in certain complex thalloid liverworts, such as Conocephalum conicum (Brown & Lemmon, 1988c) where spores are not in tetrahedral arrangement and cleavage planes are determined after meiosis by interaction of opposing sets of radial microtubules emanating from the four nuclei. Phragmoplasts initiated in these radial microtubule systems mediate intersporal septa that cleave the spores in various patterns.

Upon completion of cytokinesis the microtubules are reorganized into an extensive RMS emanating from the nucleus in each of the newly cleaved spores. The RMS in the young spores appears to be associated with positioning of the nucleus and underlies the developing sporoderm. Although it seems likely that microtubules are associated with deposition of the sporoderm, there are no direct correlations with the development of spore wall patterns. Aperture development in mosses provides the best evidence for a role of the microtubules in wall development. Immediately following cytokinesis an MTOC occurs between the nucleus and the adjacent distal wall in each young spore. The MTOC with microtubules radiating from it then moves from the distal surface to the proximal pole located at the former center of the sporocyte (Brown & Lemmon, 1983). There the MTOC and the radial system of microtubules is involved in development of the proximal aperture.

Diversity of Liverworts

The critical position of liverworts (Marchantiophytina) with regard to land plant evolution is uncontested. All recent analyses find the liverworts to be a single monophyletic group. As an early diverging lineage of embryophytes, the liverwort clade is as old or older than any other land plant lineage (Mishler & Churchill, 1984; Hedderson et al., 1996, 1998; Garbary & Renzaglia, 1998; Qiu et al., 1998; Renzaglia et al., 2000; Von Konrat et al., 2010a; b). The liverworts are an extraordinarily diverse group with numerous structures unique in the plant kingdom. Estimates of the number of liverwort species vary by as much as 50 %, with estimates ranging from 4,500 to 9,000 (Von Konrat et al., 2010a, b). Estimates of about 5,000 species in some 400 genera have been widely accepted (Crandall-Stotler et al., 2009a, b). Liverworts are distinguished by several synapomorphies: Oil bodies, the growth hormone lunularic acid rather than abscisic acid found in Charophycean algae and all other embryophytes, POs in mitosis, and elaters in sporogenesis. For such an extraordinarily diverse and evolutionary important group of organisms, they are underrepresented in terres of detailed cytological study. This is unfortunate as it seems probable that they may contain clues to the evolution of the organization and cycling of microtubules associated with cell division and development in plants.

In addition to the morphological diversity of liverworts, an astonishing diversity exists in the basic mechanisms of cell division. Most notably is the occurrence of MTOCs in the form of POs. The presence of "centrospheres" in cell division of liverworts was described by the early cytologists and quadripolar spindles in meiosis arising from four such centrospheres were clearly described prior to 1900. These early studies carefully described organelles at the center of a fine fibrillar network, termed filaroplasm, which we now know to be microtubules nucleated from the highly structured PO.

Three basic forms of gametophyte organization occur among the liverworts, commonly referred to as complex thalloid, simple thalloid, and leafy body plans. Traditional systems of liverwort classification largely mirror these three body plans. According to a recent synthesis (Crandall-Stotler & Stotler, 2000), the phylum Marchantiophyta (liverworts) is divided into three classes, the small early divergent Haplomitriopsida, Marchantiopsida (complex thalloids) and Jungermanniopsida (simple thalloid and leafy). The Haplomitriopsida is a very small (with just three genera), but evolutionarily important, group of liverworts about which we have little information on sporogenesis (Brown & Lemmon, 1986; Renzaglia et al., 1994).

Schuster (1984) postulated a basic dichotomy in liverworts based largely on variation in sporogenesis. The leafy and simple thalloid liverworts have quadrilobed sporocytes, whereas the sporocytes of complex thalloids are generally unlobed. In meiosis of liverworts with quadrilobed sporocytes, division polarity is determined early in prophase I and the sporocyte becomes deeply lobed with cleavage furrows that predict the eventual planes of cytokinesis. Unlike the monoplastidic sporocytes of mosses and hornworts in which the establishment of quadripolarity is associated with two divisions of the single plastid, the sporocytes of most liverworts are polyplastidic. It has only recently been discovered that monoplastidic meiosis does occur in isolated taxa scattered in the orders Calobryales, Blasiales, Monocleales (Renzaglia et al., 1994), and Marchantiales (Shimamura et al., 2001). These unexpected findings suggest that liverworts contain a number of relictual taxa that may well hold even more clues to the evolution of plant form.

For the size and importance of the group, information on the cytoskeleton and mechanisms of meiosis is known for relatively few, especially since the examples studied reveal an unexpected diversity in mechanisms of sporogenesis both among and within the major clades of liverworts. Such diversity involving basic cellular mechanisms such as polarity and spindle construction is quite unlike the other two groups of bryophytes (mosses and hornworts) and the other major plant groups. The outstanding feature of bryophyte sporogenesis in all groups is the establishment of division polarity in meiotic prophase.

Several distinct patterns of sporogenesis may be recognized in the liverworts. Whereas the Jungermanniopsida exhibit extreme quadrilobing, the Marchantiopsida are characterized by a gradation in quadripolarity and a diversity of spindle types and patterns of meiosis. The members of the Jungermanniopsida all have sporocyte quadripolarity that is predicted by intersecting bands of microtubules marking the sites of the future furrows. This newly recognized distinctive feature of the cytoskeleton is also found in the basal grade liverwort Blasia where it occurs in association with monoplastidy. Blasia exhibits a unique combination of premeiotic bands, deeply lobed sporocytes, and plastid polarity. Several complex thalloids exhibit monoplastidic meiosis with precocious establishment of quadripolarity similar to mosses and hornworts, but without deep cytoplasmic lobing (Shimamura et al., 2000, 2001, 2004; Brown et al., 2010). In these monoplastidic liverworts, as in other monoplastidic bryophytes, the plastids are distributed to the four future tetrad poles where they serve as MTOCs. Microtubules nucleated at the four P-MTOCs give rise to a quadripolar spindle. At the extreme are the apolar sporocytes of Conocephalum conicum (Brown & Lemmon, 1988c) and Conocephalum japonicum (Shimamura et al., 1998) and certain allied taxa (Shimamura et al., 2011) that undergo polyplastidic meiosis in a manner more typical of seed plants. Meiotic spindle microtubules are nucleated by [gamma]-tubulin in the perinuclear area (NE-MTOC) and the spindle is gradually organized into a bipolar structure (Shimamura et al., 2004). Following meiosis, RMSs emanating from the four nuclei define spore domains in the brief coenocyte and initiate phragmoplasts that mediate cytokinesis (Brown & Lemmon, 1988c; Shimamura et al., 1998). Thus, the pattern of spore cleavage is controlled by the position of tetrad nuclei after meiosis, rather than having been determined before meiosis as is more typical of bryophytes.

Next we describe features of meiosis in examples from the two major liverwort classes, the Marchantiopsida and Jungermanniopsida. Unfortunately there is no comparable data on the Haplomitriopsida, although it is known from TEM studies that both monoplastidic and polyplastidic meiosis occurs among taxa traditionally treated in the genus Haplomitrium (Renzaglia et al., 1994) suggesting that detailed investigation of cell division in this group will provide important information regarding early land plant evolution.

Complex Thalloid Liverworts

The complex thalloid liverworts (Marchantiopsida) are an early divergent and highly evolved group of organisms. Putatively beginning with a more simple general thalloid body plan like that of today's Monoclea (Schuster, 1992) or Blasia (Renzaglia et al., 2007), the Marchantiopsida radiated to give rise to forms with elaborate elevated gametangiophores and reduced forms with receptacles hidden within the thallus, reduced sporophytes and loss of elaters. Accordingly, the small order Blasiales is now considered sister to all other marchantiods and the Ricciales the most highly reduced and in many respects, the most derivative family of Marchantiales. Strong selection to both reproduce rapidly and provide maximum protection for gametangia in the Ricciaceae (Schuster, 1992) was accompanied by evolution of exceptional desiccation tolerance and ability to colonize disturbed areas ("weediness").

In the small and early divergent order Sphaerocarpales, Sphaerocarpos, seems to employ an alternate strategy of ephemeral habit for rapid colonization of waste areas. Sphaerocarpos has a rapid life history; in North America producing spores during the wet winters of the Gulf Coast or following Show melt in more northern latitudes. The only other two genera, Riella and Geothallus, are aquatics of ephemeral water bodies surviving drought as spores. For a thorough discussion of evolution of the complex thalloids see Schuster (1992) and Forrest et al. (2006).

In addition to morphological diversity, an astonishing diversity exists in meiosis. Although far from complete, we have a decent knowledge of diversity in sporogenesis in this group (Brown et al., 2010; Shimamura et al., 2011). We can envision a trend from monoplastidy to oligoplastidy to polyplastidy to apolar sporocytes.

Monoplastidy is rather widespread in this group especially among early divergent and basal members. Monoplastidic meiosis has been reported for 3 of the 5 orders of Marchantiopsida; Blasiales, Lunulariales, and Marchantiales. Blasia, which is of paramount importance because molecular phylogenetic studies show it to be sister to all other marchantioids, has a unique mechanism of meiosis in which quadripolarity of deeply lobed sporocytes is determined by both premeiotic bands and plastid polarity (Brown et al., 2010). Monoplastidic meiosis has also been demonstrated in Cavicularia, the only other genus in the Blasiales (Shimamura et al., 2005). Although Lunularia (Lunulariales) is clearly monoplastidic (Shimamura et al., 2011), we have no information on the cytoskeleton and there have been no investigations of meiosis whatsoever in the Neohodgsoniales. Monoclea gottschei, which has monoplastidic meiosis and was traditionally in its own order, is now grouped with the monoplastidic Dumorteria and Wiesnerella in the heterogeneous Marchantiales (Crandall-Stotler et. al., 2009a, b; Shimamura et al., 2011). Both M. gottschei and D. hirsuta undergo monoplastidic meiosis with precocious establishment of quadripolarity that is similar to mosses and hornworts, but without deep cytoplasmic lobing (Shimamura et al., 2000, 2001, 2004). In early meiotic prophase of D. hirsuta, as in other monoplastidic bryophytes, the plastids are distributed to the four future tetrad poles where they serve as MTOCs. It was in this organism that the ability of microtubules to assemble from isolated plastids proved experimentally that plastids serve as MTOCs in monoplastidic meiosis (Shimamura et al., 2004).

Meiosis in Marchantia can best be described as oligoplastidic and intermediate between monoplastidic meiosis in Dumortiera and the polyplastidic type of the core marchantioids in which signs of polarity are lost in the polyplastidic cytoplasm but still evident in the origin of the spindle (Brown et al., 2010). During prophase I in meiosis of Marchantia, the plastid undergoes regular division and positioning to establish the four tetrad poles, but the plastids continue to divide in the spore domains. Whereas the [gamma]-tubulin is associated with plastids initially, it becomes distributed to the tetrad poles where it occurs as diffuse assemblages that give rise to the meiotic spindle (Brown et al., 2007).

At the other extreme, Conocephalum conicum (Brown & Lemmon, 1988c) and C. japonicum (Shimamura et al., 1998) undergo polyplastidic meiosis in unpolarized sporocytes in a manner similar to seed plants. Spindle microtubules are organized by an NE-MTOC and the anastral spindle is gradually organized into a bipolar structure with no trace of quadripolarity. The spindles of both meiotic divisions are variously oriented and spores are cleaved at boundaries of radial microtubules emanating from the tetrad nuclei after meiosis. The variable pattern of spore tetrads in C. conicum was important in recognizing that nuclear cytoplasmic domains (NCDs) control division polarity in plant cell division (Brown & Lemmon, 1992, 2001). Variable patterns of spore tetrads suggest that there may be other examples of apolar sporocytes in the Marchantiopsida (Shimamura et al., 2011).

Monoplastidic Meiosis and Pre-meiotic Bands in Blasia pusilla

The quadrilobed sporocytes of B. pusilla are striking with a single large green plastid filling each of the lobes (Fig. 7). Sporocyes of Cavicularia densa, the only other species in the Blasiales, are monoplastidic and appear very similar (Shimamura et al., 2005) but unfortunately details of sporogenesis have not been studied. Immunofluorescence of microtubules reveals an unexpected combination of configurations found in the lobed polyplastidic sporocytes of Jungermanniopsida and monoplastidic sporocytes of the complex thalloids. Blasia is the only known bryophyte that has both pre-meiotic bands and monoplastidic meiosis.

There is extensive cytoplasmic preparation for quadripartitioning of the sporocyte that begins well in advance of nuclear division (Fig. 8). The early sporocyte emerges as an ovate, unlobed cell with two plastids (Fig. 8a, d). Quadrilobing is initiated by a cortical microtubule system that begins as a single band (Fig. 8a) and branches/ bifurcates to mark infurrowing regions of the cytoplasm (Fig. 8b, c). This initial infurrowing does not appear to involve plastids, which at this time are in the interior of the cytoplasm clustered around the nucleus (Fig. 8e).

The pre-meiotic bands form a cloverleaf pattern that defines infurrows and the sporocyte becomes deeply quadrilobed (Fig. 8c) with a single large plastid positioned in each of the lobes (Fig. 8f). Such pre-meiotic bands of microtubules have been reported in deeply lobed, polyplastidic sporocytes of both simple thalloid (Brown & Lemmon, 2004, 2006) and leafy (Brown & Lemmon, 2009) liverworts. It appears that the interlocking bands predict the complex cleavage planes of simultaneous cytokinesis after meiosis in much the saine way as the single PPB predicts the simple cleavage plane of mitosis. The complex system of microtubule bands in meiosis marks the planes of cytokinesis that will simultaneously separate spores of the tetrad after meiosis. The cytoskeletal configurations associated with division quadripolarity are shown in stereo in Fig. 9. The system of interlocking cortical bands of microtubules involved in the early development of quadrilobing (Fig. 9a) is replaced by the plastid based QMS that is precursor to the quadripolar spindle (Fig. 9b).

The furrow-associated rings of microtubules disappear and are replaced by a reticulum of microtubules throughout the cytoplasm of the sporocyte. Microtubules gradually become more focused on the four tetrahedrally arranged plastids, first around their entire surfaces and then concentrated on the proximal surfaces (Fig. 10a). Opposing ends of microtubules emanating from the four plastids interact to form a QMS encaging the nucleus (Fig. 10b). The poles of each array remain closely associated with the plastid envelope. The QMS becomes transformed into the functionally bipolar spindle of meiosis I by increased development of spindle fibers and movement of pairs of poles toward a single division axis (Fig. 10c). As kinetochore fibers form among interacting sets of microtubules of the QMS, astral microtubules radiating from the poles gradually disappear (Fig. 10c). The four-covered metaphase I spindle has the shape of a twisted rectangle (Fig. 10d) with each of the comers adjacent to proximal sides of the tetrahedrally arranged plastids. This spindle configuration in the shape of a twisted rectangle with boxlike poles appears to be the fundamental spindle type of marchantioids.

As the kinetochore fibers shorten in anaphase, the spindle becomes more rod-shaped (Fig. 10e). It delivers the two groups of chromosomes to positions adjacent to the polar furrows midway between pairs of plastids in anaphase and a phragmoplast is constructed in the midzone (Fig. 10f). Second division occurs in the undivided sporocyte and the completion of meiosis results in inheritance of a single plastid and chloroplast in each young spore. Cytokinesis is simultaneous.

Monoplastidic Meiosis in Monoclea gottschei

The early sporocyte emerges as a still elongate cell with the single plastid either preparing for division or just divided (Fig. 11a). The sporocyte becomes more spherical and the two daughter plastids migrate to opposite positions and undergo a second division (Fig. 11b). The four plastids become tetrahedrally positioned in the sporocyte. The nucleus is drawn out into a tetrahedron with broad lobes adjacent to the plastids (Fig. 11c). Following establishment of quadripolarity the cytoplasm become slightly lobed, the sporocyte wall definitely thicker in the cleavage furrows (Fig. 11d). The plastids may continue to divide resulting in a cluster of plastids in each of the lobes (Fig. 11d).

Microtubules emanating from the plastid poles forma QMS (Fig. 12a) surrounding the central nucleus (Fig. 12c). The nucleus becomes somewhat distended into a tetrahedron, but less so than in deeply lobed sporocytes of the Jungermanniopsida. The extent of the cytoplasmic lobing can be seen in sporocytes flattened during processing as in Fig. 12a. The localization of [gamma]-tubulin (Fig. 12b) seems less precise than in other monoplastidic sporocytes, perhaps because the plastids have proliferated at each pole and remain associated with the actual center of microtubule organization.

The spindle is quadripolar in origin (Fig. 12d-f). [gamma]-Tubulin (Fig. 12e) appears to be distributed throughout the microtubule system felting the nucleus. The resultant metaphase I spindle (Fig. 12g-i) shows little evidence of its quadripolar origin. The spindle has a broad equatorial region aligned in the equatorial furrow and sharply focused poles terminating at the polar cleavage furrows (Fig. 12g). The [gamma]-tubulin (Fig. 12h) extends along the spindle fibers.

In late anaphase/telophase the spindle is sharply focused (Fig. 13a-c). [gamma]-tubulin is clearly associated with the shorted spindle fibers in the polar regions (Fig. 13b). A spectacular system of phragmoplasts is built in inframeiotic interphase (Fig. 13d-f). The complex consists of BMAs that emanate from nuclei and flanking plastids (Fig. 13d). Similar constructions are seen in monoplastidic meiosis of mosses and hornworts as microtubules emanate from plastids as well as the proximal surfaces of nuclei. The influence of the furrows can be seen in the slightly reniform nuclei (Fig. 13f). No cell plate is deposited and meiosis II occurs in the undivided cytoplasm (Fig. 13g-i). The second division spindles are quite pointed (Fig. 13g) and deliver nuclei to the tetrad poles. Phragmoplasts (Fig. 13j-1), which develop first between sister nuclei and then between non- sister nuclei, form a complex that directs deposition of intersporal septae resulting in simultaneous cytokinesis.

Monoplastidic Meiosis in Dumortiera hirsuta

Monoplastidic meiosis in D. hirsuta has been more thoroughly studied than any other marchantiod (Shimamura et al., 2000, 2003, 2004; Brown et al., 2010). Following the two rounds of plastid division, sporocytes are only slightly lobed, but the plastids are clearly in tetrahedral arrangement (Fig. 14a-c). When seen edge-on the wing-like plastids appear to interact with the nuclear envelope (Fig. 14b). Microtubules (Fig. 14a) emanate from the P-MTOCs (Fig. 14b) to initiate a QMS around the central nucleus, which itself is distorted into a tetrahedron (Fig. 14c). As the QMS develops (Fig. 14d), the micrombules emanating from poles interact to form BMAs interconnecting the plastids (Fig. 14d). [gamma]-tubulin is concentrated at poles of the QMS in association with the proximal surfaces of the plastids, although a portion of the [gamma]-tubulin seems to localize along micrombules (Fig. 14e) outlining the tetrahedral nucleus (Fig. 14f).

In prometaphase I, the poles of the QMS converge in pairs toward the spindle axis (Fig. 14g-i). [gamma]-Tubulin is concentrated at the four poles (Fig. 14h) and the nucleus is drawn into tips at these regions (Fig. 14i). Chromosomes are becoming distinct at this stage.

The metaphase I spindle (Fig. 15a-c) has boxlike poles with a slightly expanded equatorial region. The [gamma]-tubulin has moved into the spindle (Fig. 15b). In telophase-I (Fig. 15d-f), opposing microtubule systems (Fig. 15d) emanate from [gamma]-tubulin that closely invests the nuclear envelopes (Fig. 15e). The first division phragmoplast is very well developed (Fig. 15d) but no cell plate is deposited.

In prophase II (Fig. 15g-i), a complex of BMAs (Fig. 15g) interconnects the four plastids located in the spore domains. Microtubules that enclose the nuclei initiate the prophase II spindles (Fig. 15g). [gamma]-tubulin remains in association with nuclear envelopes and punctuate staining reappears on the plastid surfaces (Fig. 15h). The metaphase II spindles (Fig. 15j-l) have pointed poles with [gamma]-tubulin (Fig. 15k) localized throughout.

Telophase II nuclei are adjacent to the plastids at tetrad poles (Fig. 16a-c) and [gamma]-tubulin invades the interzones (Fig. 16b). Following telophase-II (Fig. 16d-f), [gamma]-tubulin becomes associated with nuclear envelopes of the tetrad nuclei (Fig. 16e) and radial microtubules emanate from the nuclei to give rise to the initial bipolar arrays of the primary phragmoplasts (Fig. 16d). Microtubules interconnect all four of the tetrad nuclei giving rise to a configuration not unlike that seen in inframeiotic interphase (Fig. 15g). The resultant phragmoplast complex guides deposition of intersporal septae in simultaneous cytokinesis (Fig. 16g-i). During this stage, [gamma]-tubulin is concentrated at nuclear surfaces (Fig. 16h).

Oligoplastidic Meiosis in Marchantia polymorpha

The early sporocyte emerges with a single plastid as seen in micrographs by Shimamura et al. (2001). Early sporocytes are ovoid to pyramidal with little tendency to lobe during the establishment of quadripolarity (Fig. 17a-l). The single plastid divides and the two plastids become positioned on opposite sides of the nucleus (Fig. 17a-c). [gamma]-Tubulin is clearly associated with the dividing plastids (Fig. 17b). The plastids rotate so that the tips are in tetrahedral arrangement (Fig. 17d-f). The [gamma]-tubulin is redistributed so that it is concentrated at tips of the plastids (Fig. 17h). As prophase progresses (Fig. 17g-i) the microtubules become focused at the tips of the plastids forming a QMS interconnecting the plastids (Fig. 17g). The nucleus becomes distorted into a tetrahedron (Fig. 17i). The plastids complete division resulting in four plastids at the tetrad poles. In late prophase, the plastids continue to divide resulting in a cluster of plastids around the tetrad poles marked by concentrations of [gamma]-tubulin(Fig. 17k).

The developing spindle passes through a distinct quadipolar or even multipolar stage. This cytoskeletal configuration is depicted in three dimensions in Fig. 18. Cones of microtubules (Fig. 18a) emanate from [gamma]-tubulin (Fig. 18b) associated with plastid aggregates in the four spore domains.

The meiosis I spindle (Fig. 19a-c) exhibits broad poles with little vestiges of its quadripolar origin. The poles appear asymmetric in equatorial view as they are slightly oblique to one another. "[gamma]-tubulin marks the broad polar regions and is absent from the equatorial region (Fig. 19b). During inframeiotic interphase (Fig. 19d-f) [gamma]-tubulin is associated with nuclear envelopes (Fig. 19e) and opposing RMSs define the dyad domains, which are separated by an OB (shown in TEM, Brown et al., 2007). Nuclei divide simultaneously in the undivided cytoplasm. The spindles of meiosis II (Fig. 19g-i) are at more or less right angles to one another and are focused on poles of dispersed "[gamma]-tubulin that extends along the polar portions of the spindle (Fig. 19h).

Following telophase II (Fig. 19j-l), [gamma]-tubulin becomes distributed evenly around the tetrad nuclei (Fig. 19k) and microtubules radiate from the nuclear envelopes (Fig. 19j). Interaction of these microtubules establishes the complex of phragmoplasts that direct simultaneous cytokinesis along the pre-determined cleavage planes resulting in a tetrahedral tetrad of spores.
COPYRIGHT 2013 New York Botanical Garden
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2013 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Title Annotation:p. 178-210
Author:Brown, Roy C.; Lemmon, Betty E.
Publication:The Botanical Review
Article Type:Report
Geographic Code:1USA
Date:Jun 1, 2013
Previous Article:Biology and functional ecology of Equisetum with emphasis on the giant horsetails.
Next Article:Sporogenesis in bryophytes: patterns and diversity in meiosis.

Terms of use | Privacy policy | Copyright © 2019 Farlex, Inc. | Feedback | For webmasters