The volvocaceans are a closely related group of green flagellates that range in size and complexity from colonial forms that contain a small number of identical cells, to Volvox in which there is a division of labor between several thousand terminally differentiated somatic cells and a small number of asexual reproductive cells called gonidia. Similar cytoplasmic bridges link the cells of all volvocacean embryos, but the formation, structure and function of such bridges have been studied most extensively during Volvox carteri embryogenesis.

Each mature V. carteri gonidium produces an embryo that executes 11-12 rounds of rapid, stereotyped, synchronous cleavage divisions in the absence of any growth. Large numbers of cytoplasmic bridges are formed in each cleavage furrow as a result of incomplete cytokinesis. The bridges are tightly packed in bands that are concentric with the inner surface of the hollow embryo, and the bridge bands of all cells are linked to form a single, coherent cytoplasmic-bridge system that runs through the entire embryo and holds it together.

A fully cleaved embryo contains all of the cells that will be present in an adult of the next generation, but it is inside-out with respect to the adult configuration: its gonidia are on the outside and the flagellar ends of its somatic cells are on the inside. This awkward arrangement is corrected by inversion, in which the embryo rapidly turns itself right-side-out through a combination of cell shape changes and movements. The cytoplasmic-bridge system plays a central role in inversion by providing the physical framework against which the cells exert force to reverse the curvature of the cellular sheet. Recent studies show that a kinesin located in the cytoplasmic bridges appears to provide the critical driving force of inversion. Inversion is a defining characteristic of the volvocaceans, and all of the volvocaceans that have been tested have a gene product that is homologous to the kinesin that drives inversion in V. carteri. So we think that the mechanism of inversion is probably similar in all of them.

In all of the smaller volvocaceans, and in about half of all Volvox species, the cytoplasmic bridges disappear soon after inversion has been completed and a newly formed ECM has taken over the job of holding the organism together. However, three lineages of Volvox have independently evolved the trait of retaining cytoplasmic bridges throughout the life of the adult. And in each case this trait has coevolved with a novel form of embryonic development in which embryos begin to divide while they are still quite small, and then grow between successive divisions—all the while remaining connected to the adult somatic cells. Here we provide new information about the formation and the structure of persistent cytoplasmic bridges in one such species of Volvox, and review the evidence that leads us to believe that adult cytoplasmic bridges serve as conduits to convey nutrients from somatic cells to developing embryos.

Introduction

Late in the 17th century Antoni van Leeuwenhoek watched with fascination as the little, round green “animalcules” to which Linnaeus would later give the name “Volvox” rolled about restlessly before his simple lens.² More than 130 years would pass before microscopes were improved sufficiently that another great microbiologist, Christian Ehrenburg, would realize that each cell within such a Volvox spheroid was connected to its neighbors by “filamentous green tubes”.³ Such intercellular connections were seen in free-swimming adults of both species of Volvox then known (V. globator and V. aureus), and in 1896 Arthur Meyer used plasmolysis to demonstrate that these intercellular connectors were actually cytoplasmic extensions of the cells, indicating that the adult organisms in both of these species were syncytial.⁴ In that same paper, however, Meyer described a third species of Volvox (which he imaginatively named V. tertius) in which fine connections existed between cells of the developing embryo, but then those bridges disappeared as the embryo developed into an adult.⁴

It would turn out that the latter pattern—cytoplasmic bridges present in the embryo but not the adult—is ancestral. It is the pattern seen in all of the smaller and simpler relatives of Volvox in the family Volvocaceae (genera such as Pandorina, Eudorina and Pleodorina) as well as in about half of all species of Volvox. Persistence of bridges in the adult is a derived trait that is seen only in some species of Volvox. We will return to a discussion of that derived trait toward the end of this chapter.

A Brief Overview of Volvox carteri Development

In her 1933 paper describing embryonic development in two species of Volvox found in her native South Africa, Mary Agard Pocock first suggested that cytoplasmic bridges probably were formed as a result of incomplete cytokinesis.⁵ Over the years, that suggestion was repeated by investigators studying development in various members of the family Volvocaceae,^1,6-18 but formation of the cytoplasmic bridges has been studied systematically only in Volvox carteri embryos.^19,20 For that reason, a brief overview of embryonic development in V. carteri is called for as background for the subsequent discussion of cytoplasmic-bridge formation and function (details of V. carteri development are available elsewhere.²¹

A young adult spheroid of V. carteri consists of several thousand small, biflagellate somatic cells at the surface of a transparent sphere of glycoprotein-rich extracellular matrix and a few large asexual reproductive cells, called “gonidia”, that lie just beneath the somatic cell monolayer (fig. 1). When a gonidium is mature, it behaves like a stem cell and initiates a series of rapid, synchronous cleavage divisions that produce a new juvenile individual containing a new cohort of somatic and gonidial cells (fig. 2). The prospective gonidial and somatic cells of the new generation are set apart during cleavage by a stereotyped set of asymmetric cleavage divisions (fig. 2B). Each large cell produced by asymmetric division produces one gonidium while its smaller sister cell produces a clone of somatic cells.

Figure 1

Volvox carteri. A young adult asexual spheroid, as shown here, contains several thousand Chlamydomonas-like, biflagellate somatic cells at the surface, and 16 large gonidia (asexual-reproductive cells) just beneath the surface, of a transparent sphere (more...)

Figure 2

Scanning electron micrographs (SEMs) of cleaving V. carteri embryos in polar view (cleavage normally takes place inside the adult spheroid.). Each mature gonidium undergoes a stereotyped set of rapid, synchronous cleavage divisions. A) The first five (more...)

By the end of cleavage the embryo is a hollow sphere in which all of the cells of both types that will be present in an adult of the next generation are already present, but the embryo is inside-out with respect to the adult configuration: its gonidia are on the outside, and the flagellar ends of its somatic cells face the interior. That awkward situation is rapidly corrected as the embryo turns itself right-side-out in a process called inversion (fig. 3). Inversion is a diagnostic feature of the family Volvocaceae; it is seen in all members of the family, and will be discussed in more detail shortly.

Figure 3

Inversion of a V. carteri embryo; A) polar view; B-D) equatorial views. A,B) Shortly after the end of cleavage the four lips of cells flanking the phialopore curl outward and backward over the rest of the embryo. C) More and more of the cellular monolayer (more...)

The Formation of a Cytoplasmic-Bridge System in Volvox carteri Embryos

Volvocacean embryos are held together during cleavage and inversion by a network of cytoplasmic bridges. Some early papers suggesting that such bridges were produced by incomplete cytokinesis^1,5-18 might have left the impression that a bridge or two was formed in each cleavage furrow. However, scanning electron microscope (SEM) examination of V. carteri embryos¹⁹ revealed that bridges are actually formed by the hundreds in each of the first few cleavage furrows (fig. 4A). Then at each division all preexisting bridges are preserved and divided between the daughter cells while many additional bridges are being formed in the furrow between them. Thus, for example, each cell of an 8-cell embryo is linked to its neighbors by ˜250 bridges, half of which were inherited, and half of which were newly formed during the third division. Bridge transmission and formation continues throughout cleavage, resulting in a progressive increase in the total number of bridges in the embryo—although the number of bridges per cell decreases, as the cells become smaller. By the end of cleavage the embryo contains about 100,000 bridges, and an average cell is connected to its neighbors by about 25 bridges.¹⁹

Figure 4

Scanning electron micrographs (SEMs) of fragments of V. carteri embryos that were dissected after they had been fixed and dried. A) One half of a 4-cell embryo. Broken cytoplasmic bridges appear as bumps (or sometimes shallow pits) on the surface of the (more...)

It is important to note that these cytoplasmic bridges are not located randomly on the cells. They are organized into closely packed, curved bands of bridges (fig. 4B) that are concentric with the inner surface of the hollow embryo, closer to the end of the cell with the nucleus and basal body than to the end of the cell containing the chloroplast (fig. 4A). Transmission electron microscopy (TEM) of thin sections reveals that throughout cleavage, the bridge bands are present at a level that is occupied on the cellular interior by a bowl-shaped zone of Golgi-rich cytoplasm that cups the nucleus and separates it from the chloroplast.²⁰ Most importantly, the bridge bands of all cells are aligned and connected, forming a single, coherent cytoplasmic-bridge system that is continuous throughout the embryo, and that holds it together.^19,20

The position of the cytoplasmic-bridge bands on the dividing cells is linked to the mechanism by which they are formed during cytokinesis.²⁰ As alluded to in the preceding paragraph, all cells of the cleaving embryo have a very regular, polarized organization: the basal bodies (BBs) are located at the inner tip of each cell, the nucleus is positioned just below the BBs and is surrounded by cytoplasm containing numerous Golgi bodies and vesicles. The large chloroplast fills the outer portion of the cell. Each cell is girdled by an extensive array of cortical microtubules (MTs) that originate near the basal bodies and extend toward the chloroplast end of the cell just under the plasma membrane. As sister nuclei are formed during mitotic telophase, parallel sheets of additional MTs emanating from the vicinity of the (now-separated) sister BBs extend through the cytoplasm in the inter-nuclear region. These parallel MT arrays are called the cleavage MTs, and they are part of the “phycoplast”, a characteristic cytoskel array that delineates the plane of cell division in most green algal cells. Division of the BB/nuclear end of the cell occurs by an ingressive furrow that is similar to the furrow often seen in metazoan cytokinesis. However, cytokinesis in the sub-nuclear region of the cell occurs by a different mechanism: by MT-guided alignment and fusion of Golgi-derived vesicles—much like the formation of a cell plate in higher-plant cytokinesis.²⁰ The cytoplasmic bridges are apparently formed at the regularly spaced places where these vesicles fail to fuse, and the bridges remain in this sub-nuclear region, closely associated with the nearby MT arrays, throughout the rest of the cleavage period.

Structural Features of the V. Carteri Cytoplasmic-Bridge System

The cytoplasmic bridges of the V. carteri embryo are extremely regular in appearance. They are ˜200 nm in diameter, spaced 500 nm center-to-center on average (fig. 4B), and are associated with at least two types of conspicuous membrane specializations. The inner face of the plasma membrane throughout the bridge region is coated with electron-dense material that appears to be thickest in the center of each bridge, so that the bridges appear as dark rings when sectioned transversely (fig. 5A,D). Intimately associated with (or embedded in) that electron-dense coating is an extensive array of concentric or spiral cortical striations with a spacing of ˜25 nm that not only girdle each bridge throughout its length (fig. 5B), they continue out under the adjacent plasma membrane of the cell body as concentric rings that abut the rings surrounding neighboring bridges (fig. 5A). In bridges that have been sectioned longitudinally, these cortical striations often appear as a series of short, uniformly spaced parallel line segments, near the cell surface (fig. 5C). Extremely similar cortical striations in the bridge region were first observed by Jeremy Pickett-Heaps in V. tertius embryos,¹¹ and were later seen in the cytoplasmic bridges of Eudorina elegans embryos.¹³ The molecular nature of the cortical striations is unknown, but there is some evidence supporting the notion that they serve to strengthen the membrane in the bridge region.²⁰ Of particular significance with respect to the role that the cytoplasmic bridges will play during inversion is the fact (mentioned earlier) that throughout embryogenesis the bridges are associated with numerous cortical MTs that originate in the BB region and run to the opposite end of the cell (fig. 5D). During cleavage, additional MTs are often seen running into or through cytoplasmic bridges, at an angle to the main MT arrays, but such MTs are no longer seen during inversion. However, a second type of cytoskeletal element is seen in association with the bridges during inversion —a set of fibers of unknown composition that connect neighboring bridges (fig. 5E). It has been postulated that these fibers account for the observation that the center-to-center spacing of the bridges is essentially constant during inversion, despite the considerable stresses that the bridge system appears to experience.²⁰ Although strands of ER have been seen within the cytoplasmic bridges of other species of Volvox,^1,7,9,11 none have been seen in the cytoplasmic bridges of V. carteri embryos.²¹

Figure 5

Transmission electron micrographs (TEMs) of cytoplasmic bridges in cleaving V. carteri embryos. A) In a thick section parallel to a cleavage furrow, each cytoplasmic bridge is cut transversely and appears as a dark ring. Note the concentric cortical striations (more...)

The Function of the Cytoplasmic-Bridge System in Cleaving Embryos

The most obvious function of the cytoplasmic-bridge system in a volvocacean embryo is to hold the embryo together. The best evidence that they actually do this is as follows: the cytoplasmic bridges of V. carteri embryos routinely break down shortly after inversion—but normally only after enough extracellular matrix (ECM) has been assembled to hold the cells together once the bridges are gone. However, if assembly of the ECM is blocked with a specific inhibitor, the bridges break down on schedule, and the juvenile then falls apart into a suspension of single cells.²²

It has been postulated that the cytoplasmic-bridge system also provides the communication channels for signals that synchronize mitotic cycling in the cleaving embryo. Mitotic activities are so well synchronized in V. carteri that, for example, when one cell is in anaphase (which lasts only five minutes), all cells are in anaphase.¹⁹ However, as appealing as it is to propose that such perfect mitotic synchrony is mediated by chemical signals passing through the cytoplasmic bridges, there is no direct evidence that this is the case.

The Function of the Cytoplasmic-Bridge System in Inverting Embryos

The role of the cytoplasmic-bridge system in synchronizing mitosis may be in doubt, but there can be little doubt that the bridge system plays a central role in inversion.^15,17,20 The bridge system that is formed during cleavage, and in which all the bridges of all cells are aligned and linked into a single, coherent network, remains intact during inversion. It constitutes the only structural framework against which the cells can exert force in order to turn the embryo inside out. And they do indeed exert force against it.

By the end of cleavage the cytoplasmic-bridge system links the cells at their widest points, just below the level of their nuclei (fig. 6A). The only place where the neighboring cells are not linked by the cytoplasmic-bridge system in this way is across the phialopore, an opening at the anterior end of the embryo that is formed in very early cleavage, and that is the site where inversion begins. The outward curling of the cellular monolayer begins when cells near the phialopore execute two coordinated activities: (i) they change shape, forming long, narrow, MT-reinforced “stalks” at their outer (chloroplast) ends and (ii) they move inward individually relative to the cytoplasmic-bridge system, going as far as they can go in that direction. In combination, these two changes cause the cells to go from being linked at their widest points in the sub-nuclear region to being linked at their slender outermost tips. And that, in turn, forces the cellular monolayer to curl outward (fig. 6B).

Figure 6

Diagrammatic representation of the cellular mechanism of inversion as deduced from EM analysis of inverting embryos. A) Before inversion the cytoplasmic-bridge system (red line) links the cells (green outlines) at their widest points, just below the nuclei (more...)

Subsequently the region of maximum curvature moves toward the posterior pole of the embryo as cells that have passed through the region of curvature withdraw their stalks and take on the appearance of a simple columnar epithelium, while other cells that were located further from the phialopore move into the region of curvature by changing shape and moving relative to the cytoplasmic-bridge system—just as the cells closer to the phialopore did earlier.

By this process, the embryo turns itself right-side-out in about 40 minutes, the BB/flagellar ends of all cells are brought to the outer surface, and all of the cells are now linked by the cytoplasmic-bridge system at their (now-innermost) chloroplast ends.^15,17,20 Shortly after inversion a primary layer of ECM is formed that encircles the embryo and holds all of the cells in a fixed orientation,²² and then the cytoplasmic bridges disappear by a process that has yet to be analyzed.

Inversion is a hallmark of the family Volvocaceae. But how general is the cellular mechanism of inversion that has been described for V. carteri? The first EM-level examination of inversion was performed by Jeremy Pickett-Heaps, studying V. tertius.¹¹ He reported that although cells in cleaving embryos had cytoplasmic bridges at the sub-nuclear level, by the end of inversion they were connected only at the chloroplast ends. He inferred (but provided no evidence) that the cells were initially connected at both the sub-nuclear and chloroplast regions, but that during inversion the sub-nuclear bridges were broken, leaving the bridges at the chloroplast ends of the cells to serve as “hinges” that would allow the cells to swing outward. Subsequently, investigators studying Pandorina morum¹⁴ and Eudorina elegans^13,16 made observations and drew conclusions very similar to those of Pickett-Heaps. However, none of the micrographs in any of those four reports shows cells linked simultaneously in the sub-nuclear and chloroplast regions: the only bridges shown in cleaving embryos are sub-nuclear, and bridges at the chloroplast ends of the cells are shown only in post-inversion cells.^11,13,14,16 Thus, all published micrographs are consistent with the concept that in P. morum, E. elegans and V. tertius—as in V. carteri—each cell possesses a single band of cytoplasmic bridges that is located in the sub-nuclear region during cleavage, but that ends up at the chloroplast end of the cells by the end of inversion. This, in turn, is consistent with the idea that the structure of the cytoplasmic-bridge system, and its function during inversion, is similar in all volvocacean embryos.

An Inversion Motor in the Cytoplasmic Bridges

The development of a transposon-tagging system for V. carteri ²³ provided an effective way of cloning developmentally important genes.^24,25 The first inversion-specific gene tagged with this transposon (which was named Jordan because it jumped so well) was invA.²⁶ In mutants in which invA has been inactivated by a Jordan insertion, inversion appears to start in the usual way, but then it stops abruptly as soon as the lips of cells flanking the phialopore have curled outward partway (fig. 7A). Although the mutant cells change shape as wild-type cells do during inversion, they fail to move with respect to the cytoplasmic-bridge system (fig. 7B). This mobility defect and the resulting inversion block can both be cured in either of two ways: (i) by excision of Jordan from the invA locus, or (ii) by transformation of the mutant strain with a wild-type invA transgene.²⁶ These findings led to the conclusion that embryonic cells require InvA (the protein product of the invA gene) to move with respect to the cytoplasmic-bridge system, in order to complete the inversion process.

Figure 7

Inversion arrest in an invA^- mutant embryo. A) The mutant appears to begin inverting quite normally, but it then stops abruptly after the lips of cells bordering the phialopore have just begun to curl backward. B) In a closer view of the cells circled (more...)

Sequencing revealed that InvA was a novel type of kinesin. Although the amino acid sequence of its motor domain is rather similar to that of the KIF4/chromokinesin subfamily of kinesins, outside of the motor domain it has no significant similarity to any of the hundreds of kinesins from other organisms that have been characterized to date (with the exception of an InvA orthologue that we have cloned from a closely related alga, as discussed below).

Immunolocalization studies revealed that InvA is located in the cytoplasmic-bridge region throughout embryogenesis (fig. 8,fig. 9). Its sequence indicated that it is most likely a plus-end-directed microtubule motor,²⁶ and the numerous cortical MTs that run past each cytoplasmic bridge (fig. 5D) are known to be oriented with their minus ends proximal to the basal bodies and their plus ends toward the opposite end of the cell. These facts led to the following simple model for the way that InvA functions during inversion: When InvA is activated during inversion it attempts to “walk” along the adjacent cortical MTs toward their distal ends. But InvA is not free to move, because it is anchored in the cytoplasmic bridges. Therefore, the force it exerts on the MTs results in the MTs—and with them the entire cell—moving in the opposite direction, past the cytoplasmic-bridge system (see fig. 10), thereby generating the curvature that turns the cell sheet inside out (as shown in fig. 6B).

Figure 8

Immunofluorescent localization of the InvA kinesin. Tubulin is labeled green, InvA red, and DNA blue. Each arrowhead marks the inner end of a cell whose outermost tip is marked with a small arrow. A) In a preinversion embryo, InvA is located in a narrow (more...)

Figure 9

Immunogold localization of the InvA kinesin at the EM level. A section through the outermost tips of three cells that are connected by four cytoplasmic bridges (outlined with dashed lines) in the bend region of an immunogold-labeled embryo. Arrows point (more...)

Figure 10

A model indicating how InvA is thought to drive inversion. InvA is a plus-end-directed MT motor that is firmly attached to the cytoplasmic bridges, which are all connected in turn to the rest of the cytoplasmic-bridge system. When InvA is activated during (more...)

Although no kinesins orthologous to InvA have been found outside of the green algae, an invA orthologue was cloned from Chlamydomonas reinhardtii, the closest unicellular relative of V. carteri.²⁶ It encodes a kinesin that is 90% identical to InvA in the motor domain, and 82% identical to it overall. More recently we have used an affinity-purified anti-InvA antibody to probe Western blots, and have detected a cross-reacting protein of the same size as InvA in protein extracts of Gonium pectorale, Pandorina morum, Eudorina elegans and Pleodorina californica, all of which are volvocaceans that are intermediate between Chlamydomonas and Volvox in size and complexity (I. Nishii, to be published). Our working hypothesis is that these InvA homologues all play a similar role in the inversion of the corresponding embryos. Attempts to determine the function of the InvA orthologue in unicellular Chlamydomonas are in progress.

The Evolutionary Origins of Cytoplasmic Bridges that Persist in the Adult

As noted earlier, in about half of the species of Volvox—and in all of the smaller and simpler volvocaceans—the cytoplasmic bridges that are present during embryonic stages break down after inversion, and are absent from adults. That clearly is the ancestral pattern. However, in certain species of Volvox the cytoplasmic bridges persist throughout the life of the adult. Smith believed that such bridges were so significant that in his 1944 taxonomic treatment of the genus Volvox he proposed that the genus be subdivided into four sections that differed with respect to the nature of the cytoplasmic bridges in the adult.²⁷ In the section Merillosphaera he placed eight Volvox species (including V. carteri) that lack any bridges in the adult. The section Copelandosphaera was reserved for V. dissipatrix, in which adult cells are connected by “delicate cytoplasmic strands smaller in diameter than flagella”. The section Janetosphaera was reserved for V. aureus, in which adult cells are connected by “cytoplasmic strands approximately the same diameter as flagella”. The remaining eight species, in which the cytoplasmic bridges in adults are so broad that the cells appear star-shaped when viewed from their flagellar ends, were placed by Smith in the section Euvolvox.

Smith's decision to sort Volvox species with persistent cytoplasmic bridges into three separate sections of the genus has been validated recently by a molecular phylogenetic analysis of 59 volvocine algae,²⁸ which indicates that the retention of cytoplasmic bridges in the adult has evolved independently in each of those three sections of the genus (fig. 11).

Figure 11

Phylogenetic relationships among 33 taxa of volvocine algae. Adapted (with simplifications approved by the author) from the minimum-evolution tree derived by Nozaki from the sequences of five chloroplast genes. Symbols to the right of the species names: (more...)

Structure and Formation of Persistent Cytoplasmic Bridges in Volvox, Section Euvolvox

Nearly a century and a half after Eherenburg discovered the cytoplasmic bridges of V. globator, Ikushima and Maruyama first examined them with the EM,^* and found that the cytoplasmic bridges of embryos and adults were extremely different in number, size and structure.⁹ In cleaving embryos, the bridges were about 100 nm in diameter and so numerous that multiple bridges were seen in every thin section. In contrast, the adult bridges were much greater in diameter, but much less numerous. Each adult cell had only four to six bridges connecting it to its neighbors, but those bridges were so broad that they made the cells appear star-shaped. Of particular interest was the structure of the “granule” that had long been known to be present in the cytoplasmic bridges of all members of the section Euvolvox, mid-way between the cells connected by that bridge.^4,5 The granule, they discovered, was actually a disk-like structure that was different from anything seen in the embryonic bridges of any other volvocacean species or in the adult cytoplasmic bridges of V. aureus (section Janetosphaera—the only other species of Volvox with adult cytoplasmic bridges that had been examined).^1,7 They named this structure the “medial body”. Each medial body was ˜200 nm thick, and consisted of two 40-nm-thick electron-dense discs separated by a less-dense region ˜120 nm thick. Although the medial bodies in their published micrographs ranged in diameter from about 300 to 1,000 nm, each of them fully spanned the region of the bridge in which it was located. In the surprisingly common cases in which two or three medial bodies were located edge-to-edge, the cytoplasmic bridge was split into that number of branches on each side of the medial body region, with one medial body spanning each branch of the bridge. At first glance the medial bodies appeared to occlude the bridges, but on closer examination it was found that they were traversed by numerous “canaliculi”⁹ (probably ER tubules).

This study raised—but did not answer—the question of how the many thin cytoplasmic bridges of the embryo might be related developmentally to the much smaller number of broad bridges that connect adult somatic cells. Nor did it provide any information about how the medial bodies are formed, or about the structure of the cytoplasmic bridges that are known to connect adult somatic cells to developing embryos in the section Euvolvox.⁵ One of us (Hoops) made some observations relative to those issues during an ultrastructural analysis of the flagellar apparatus in another species in the section Euvolvox, V. rousseletii,²⁹ which had previously been the object of several developmental studies.^5,30,31 His previously unpublished observations regarding V. rousseletii cytoplasmic bridges will be reviewed next.

There are at least three types of cytoplasmic bridges in V. rousseletii: “E-E” bridges linking embryonic cells to each other, “S-S” bridges linking adult somatic cells to each other, and “S-E” bridges linking adult somatic cells to developing embryos (fig. 12).

Figure 12

Three types of bridges in an adult spheroid of V. rousseletii that contains a developing embryo. S-S, a bridge connecting two somatic cells of the adult; S-E, a bridge connecting an adult somatic cell to a cell of the embryo; E-E, bridges connecting cells (more...)

In certain respects, the E-E bridges of V. rousseletii resemble those of V. carteri embryos that were discussed above: they are about 170 nm in diameter, occur in regularly spaced clusters (fig. 12, fig. 13A,B) that are located at the sub-nuclear level in cleaving embryos (fig. 12, upper set of E-E bridges), but at the chloroplast ends of the cells by the time inversion has been completed (fig. 13A). They differ from the V. carteri bridges in one important respect, however: an ER tubule traverses each V. rousseletii E-E bridge (fig. 13A,B), whereas ER has not yet been detected in any of the cytoplasmic bridges of V. carteri embryos.²⁰ (We note with great interest, however, that ER tubules also traverse the cytoplasmic bridges in V. aureus,^1,7 a member of the section Janetosphaera, in which persistent cytoplasmic bridges apparently evolved independently).

Figure 13

Cytoplasmic bridges in V. rousseletii embryos and juveniles. A) Longitudinal section through bridges connecting cells of a recently inverted embryo. Note the ribosome-studded ER (arrow) traversing at least two of these bridges. B) Transverse sections (more...)

During embryonic stages, ribosomes are generally absent from much of the ER tubule that is located within a bridge (fig. 13A, arrow), even though these same tubules are studded heavily with ribosomes in the cell body. Moreover, a little later—by the time the spheroid is beginning to accumulate extracellular matrix and expand—ribosomes are entirely absent from the part of the ER tubule located within the bridge (fig. 13C). At this time the ER tubule within the bridge also becomes markedly swollen, and electron-dense material that is different from anything seen in the cell body begins to accumulate between the ER membrane and the bridge plasma membrane (fig. 13D). We believe that this accumulation of electron-dense material in the bridge represents an early step in medial body formation, but so far we have not observed other intermediate stages of medial body construction. Nor have we detected any differences between S-S bridges and prospective S-E bridges during these early stages of medial body formation.

The medial bodies in S-S bridges between two adult somatic cells are symmetrical, with two flat ˜40 nm-thick electron dense disks separated by a less dense disc of about the same thickness (fig. 14A). They range fairly widely on either side of the average diameter of ˜250 nm, but (as in V. globator) whatever the diameter of a medial body, it appears to completely span the part of the cytoplasmic bridge where it is located. ER elements continue to traverse the bridges after the medial bodies have appeared (fig.14A, arrows). In a longitudinal section through the edge of a bridge, some of those ER elements are seen to be branched and convoluted (fig. 14B, arrow), and in transverse sections (not shown) the ER elements are often seen to be arranged in a ring near the bridge periphery. Whereas the ER elements within the bridges appeared to be dilated during the early-post-embryonic stage, by this time the portions within the bridge are narrower, appearing to be pinched by the medial body.

Figure 14

Cytoplasmic bridges in V. rousseletii young adults (early in the period of ECM deposition and spheroid expansion). A) an S-S bridge connecting two young adult somatic cells. A symmetrical medial body, consisting of two similar electron-dense layers separated (more...)

The medial bodies of S-E bridges also have two electron dense layers surrounding a more electron lucent layer (fig.14C,D), but they are hourglass-shaped rather than disk-shaped as are the medial bodies between somatic cells. The minimum diameters of the S-E medial bodies are only slightly larger than the S-S medial bodies. However, the plate on the side towards the somatic cell is convex and both it and the electron lucent layer are slightly thicker than the comparable structures between somatic cells. More importantly, the dense structure on the side toward the developing embryo is about five times as thick as a disk in an S-S bridge (and about 4 times thicker than the disk on the somatic cell side of the same medial body). It has a highly crenulated outer edge (fig. 14C,D). Thus the S-E medial body is highly asymmetric, unlike the S-S medial body. Similar numbers of ER elements are seen on both sides of the medial body in an S-E bridge (fig. 14C,D, arrows), and we believe that some or all of these ER strands extend through it. However, owing to the thickness and opacity of the medial body, we have not been able to establish with certainty that they do.

When either S-S or S-E medial bodies are initially formed, and the cell bodies are still close together, the cytoplasmic bridges are widest next to the cell bodies and taper from both sides toward the medial body, which lies in the narrowest part of the bridge at that stage (fig. 14). However, as adjacent cell bodies are moved apart by the process of ECM deposition and spheroid expansion, the cytoplasmic bridges become drawn out into increasingly narrow strands, leaving the medial body region to become the widest part of the bridge.^5,30 As in V. globator, whenever two V. rousseletii medial bodies are found side-by-side, the bridge is bifurcated in that region (fig. 14C).

What Is the Relationship between Embryonic and Adult Cytoplasmic Bridges?

The above studies do not address the question of how the four to six broad bridges that are present in V. rousseletii adult somatic cells are related developmentally to the many, thin cytoplasmic bridges present in an embryo. Because there is no evidence that new cytoplasmic bridges can be formed in any volvocacean after cleavage and inversion, we believe that the medial body-containing bridges of the adult almost certainly develop from the more numerous, but smaller bridges found in the embryo. How might this happen? Two possibilities suggest themselves: either a few of the embryonic bridges might enlarge and become transformed into adult bridges while the rest of the bridges break down, or each adult bridge might be formed by fusion of several embryonic bridges (followed by additional changes in structure, of course).

We strongly prefer the latter possibility. Among our reasons are the following: (i) Whereas each bridge in the embryo is traversed by a single ER tubule (fig.13), each adult bridge is traversed by multiple ER tubules (fig. 14). We find it easier to visualize how multiple ER elements traversing each bridge might result from side-to-side fusion of several small bridges that had one tubule each than by the de novo development of multiple tubules within an existing bridge. (ii) Pocock⁵ was the first to observe that in V. rousseletii adults many cytoplasmic bridges are seen that branch in the middle and connect one cell to two others. Such Y-shaped bridges have since been documented in all published light micrographs that show an adult Euvolvox spheroid at a magnification and focal plane such that numerous adult bridges are visible.^4,5,9,30,32 A Y-shaped bridge could not be formed by enlargement of a single embryonic bridge, because each embryonic bridge is unbranched and always connects only two cells—sister cells. However, a branching bridge could be formed by fusion of two bridges that were located near one another on one cell, a, but that connected cell a to two different sister cells, b, and c (because those bridges had been formed in two different division cycles). We believe that the regular presence of Y-shaped bridges in Euvolvox adults provides the strongest evidence supporting the hypothesis that adult cytoplasmic bridges are formed by fusion of embryonic bridges.

A very different kind of support for the notion that such a fusion of bridges might be possible comes from the observation that treatment of a fully cleaved V. carteri embryo with concanavalin A can cause adjacent embryonic bridges to fuse.²¹

The Functions of Persistent Cytoplasmic Bridges

In the 1830s, Ehrenburg considered Volvox globator to be a “social animal” composed of thousands of sentient individuals that cooperate to establish and achieve common goals—such as swimming toward the light.³ When he discovered that the individual “animals” (now known as somatic cells) of a V. globator spheroid were all interconnected by “filamentous tubes” (a.k.a., cytoplasmic bridges) he had no hesitation about assigning those tubes a function. To him it seemed obvious that they must be the channels by which the individual animals communicated with each other to set and achieve their common goals. The idea that the cytoplasmic bridges probably functioned for intercellular communication and coordination of phototactic behavior persisted for decades after V. globator had come to be thought of as a multicellular alga, rather than as a social animal. But Meyer cast serious doubt on this idea when he reported in 1896 that Volvox tertius adults lack any intercellular connections, but are just as well coordinated, and swim toward the light just as well as V. globator adults do.⁴ A decade later Mast used microsurgery to show that even in species such as V. globator in which the cytoplasmic bridges persist in the adult, those bridges are not required for phototactic coordination.³³ No evidence inconsistent with Mast's conclusion has ever been published. Nevertheless, the notion that the bridges might serve for communication and phototactic coordination continues to raise its head from time to time (e.g., see ref. 9).

Another idea that has surfaced occasionally is that persistent bridges are required to hold adult Euvolvox cells in fixed positions. Pocock provided evidence that this is not the case: she reported that the bridges in V. rousseletii adults tended to break down under adverse environmental conditions, but that a spheroid remained intact even after all cytoplasmic bridges had disappeared from an entire quadrant of it.⁵ Clearly, in the section Euvolvox, as in the section Merillosphaera, once an ECM has been assembled it is able to hold the cells in place in the absence of cytoplasmic bridges.

The third function for persistent bridges that has been proposed—and the one that we find most attractive—is that they serve to channel materials from the somatic cells to the developing embryos. This idea was first proposed by Ferdinand Cohn in 1875 in one of the most important papers on Volvox development of the 19th century.³⁴ Much more recently it has reappeared as one form of the source–sink hypothesis that Bell proposed to account for the evolution of sterile somatic cells in Volvox.³⁵ If such an asymmetric flow of materials from somatic to germ cells were proven to take place through the cytoplasmic bridges in members of the section Euvolvox, it would be analogous (but clearly not homologous) to the asymmetric flow of materials through plasmodesmata in higher plants (see ref. 36 for a review), or the flow of materials from nurse cells to developing oocytes in Drosophila and other dipterans (see ref. 37 for a review).

In evaluating this material-flow hypothesis, it is important to consider the following important fact: In all three of the Volvox lineages in which persistent cytoplasmic bridges have evolved independently (fig. 11), they have coevolved with a novel form of asexual reproduction. In all species of Volvox that lack bridges in the adult—plus all of the smaller volvocaceans, and hundreds of other species of green algae—asexual reproduction involves “multiple fission”, rather than the more familiar binary fission. In multiple fission, an asexual reproductive cell (or “gonidium”) first grows 2n-fold and then executes n rapid division in the absence of any further growth, thereby forming 2ⁿ daughter cells.²¹ This clearly is the ancestral pattern of cell division in the volvocine algae. However, in all three sections of the genus Volvox in which cytoplasmic bridges are present in the adult, a different pattern is observed: Gonidia in species of Volvox with persistent bridges begin to divide while they are not a great deal larger than somatic cells, and then they grow before they divide again, nearly doubling in mass between successive divisions. Although this resembles the pattern of growth and development characteristic of the land plants and animals with which most people are more familiar, it clearly is a novel, evolutionarily derived trait in the volvocine algae. And—it bears repeating—this same deviation from the ancestral pattern appears to have coevolved with persistent cytoplasmic bridges in three independent lineages. We interpret this to mean that the two phenomena are linked in terms of their selective value: that is, that the new pattern of embryonic growth provides a selective advantage only in the pre sence of persistent cytoplasmic bridges through which materials can flow, and vice versa.

The asymmetric structure of the medial body in an S-E bridge (fig. 14C,D) is consistent with the hypothesis that the bridges and/or their medial bodies may function asymmetrically, permitting unidirectional flow of nutrients.

The rate of growth of a Euvolvox embryo can be prodigious; it can nearly double its mass, and then divide, every hour for 12-15 hours. This is a growth rate considerably greater than even lab strains of yeast can achieve, and it may well be a eukaryotic record worthy of Guinness-Book recognition. Considering this rather astonishing growth rate, Ferdinand Cohn stated matter-of-factly, with respect to V. globator, that “…it becomes obvious that the substances produced in the sum of the somatic cells during their life (carbohydrates, protoplasm, chlorophyll) aid the eight reproductive cells, so that the young colonies are not exclusively nurtured by the mother cells, but also by the united efforts of the whole cell family”.³⁴ Mary Agard Pocock, in contrast, expressed the opposite opinion—that in V. rousseletii, “…the bulk of the necessary food, which must be considerable, is manufactured by the daughter [the embryo] itself”. However—astute observer and objective scientist that she was—she went on to provide what remains about the best evidence indicating that the somatic cells do export materials to the developing embryos through the cytoplasmic bridges. She noted that that the somatic cells immediately surrounding a developing embryo usually are “nearly devoid of starch”, which she took to indicate that at least some of the nutrient required for growth of the embryo came to it from somatic cells, through the cytoplasmic bridges. She also pointed out that V. rousseletii somatic cells always grow until the gonidia begin to divide, at which time they stop growing and then later decrease in size, “particularly in the posterior region” of the spheroid where the developing embryos are located.⁵ She also noted that a V. rousseletii somatic cell that had suffered accidental severance of all of its cytoplasmic bridges was larger than its neighbors, and suggested that this probably was because it was unable to export materials through the cytoplasmic-bridge system—thereby implying once more that somatic cells normally do export such materials.

Altogether, then, there is a fair amount of indirect evidence compatible with the hypothesis that in Volvox species that have persistent cytoplasmic bridges, material may flow through the bridges from somatic cells (“the source”) to gonidia and/or embryos (“the sink”).³⁵ But as yet there is no direct evidence that this is the case. However, with all of the methods that have been developed for labeling various cytoplasmic components and monitoring their subsequent distribution, direct tests of this hypothesis should now be feasible.

Assuming that material does flow from somatic cells to embryos, what path would it be most likely to take through the cytoplasmic bridges? We assume that the ER would be involved, because adult cytoplasmic bridges that evolved independently in two sections of the genus Volvox—the sections Janetosphaera and Euvolvox—are both traversed by ER (unfortunately, EM studies of the cytoplasmic bridges in the section Copelandosphaera have not been reported to date.). One possibility is that materials are moved in the space between the ER and the plasma membrane by ER-associated molecular motors, as in the movement of certain cargos through plasmodesmata (reviewed in ref. 36). In the micrographs in Figure 14, the space between the ER and the plasma membrane is filled with electron-dense material that would appear to be a barrier to transport; but we have no way to rule out the possibility that this material is more permeable than it appears. It may even be that its apparent density is at least partially a fixation artifact (much as the often-observed “neck constriction” in plasmodesmata is thought to be at least partially an artifact of tissue preparation and fixation).³⁸ So it is possible that the medial body is more open in living cells than it appears to be in electron micrographs. Nevertheless, we believe that the morphology of the medial bodies favors the notion that material flow occurs via the ER lumen. Higher plants provide a precedent for such transport, because although much of the transport through plasmodesmata is believed to occur outside the ER (as mentioned above), there are some cases where the movement of materials appears to be through the “dilated desmotubules” (modified ER) instead.³⁶

If the ER lumen were to provide the conduits for material flow, what role might the medial bodies play? We note the lack of precedent for motor proteins within the ER cisternae and wonder if the medial bodies might supply contractile activity from the exterior, causing something resembling peristaltic waves in the ER? As we noted earlier, in fixed specimens the ER tubules of adult bridges look as though they are being pinched by the medial bodies; but in life might such pinching be a dynamic process, promoting unidirectional flow through the ER lumen? Immunocytology of medial bodies with antibodies directed against a variety of cytoskeletal and motor proteins could provide a step toward learning whether something of that sort might be possible.

References

1.

Bisalputra T, Stein JR. The development of cytoplasmic bridges in Volvox aureus. Can J Bot. 1966;44:1697–1702.

2.

van LeeuwenhoekA. Concerning the worms in sheeps, livers, gnats and animalicula in the excrement of frogs. Phil Trans R Soc Lond. 1700;22:509–518.

3.

Ehrenberg CG. Zur Erkenntniss der Organisation in der Richtung des kleinsten Raumes. Abhandl Kgl Akad Wiss Berlin. 1832:1–154.

4.

Meyer A. Die Plasmaverbindung und die Membranen von Volvox. Bot Zeitg. 1896;54:187–217.

5.

Pocock MA. Volvox in South Africa. Ann South Afr Mus. 1933;16:523–625.

6.

Metzner J. A morphological and cytological study of a new form of Volvox-I. Bull Torrey Bot Club. 1945;72:86–113.

7.

Dolzmann R, Dolzmann P. Untersuchungen über die Feinstruktur und die Funktion der Plasmodesmen von Volvox aureus. Planta. 1964;61:332–345.

8.

Stein JR. On cytoplasmic strands in Gonium pectorale (Volvocales). J Phycol. 1965;1:1–5.

9.

Ikushima N, Maruyama S. Protoplasmic connections in Volvox. J Protozool. 1968;15:136–140.

10.

Deason TR, Darden JrWH, Ely S. The development of the sperm packets of the M5 strain of Volvox aureus. J Ultrastruct Res. 1969;26:85–94. [PubMed: 5775376]

11.

Pickett-Heaps JD. Some ultrastructural features of Volvox, with particular reference to the phenomenon of inversion. Planta. 1970;90:174–190. [PubMed: 24500745]

12.

Marchant HJ. Plasmodesmata in algae and fungiIn: Gunning BES, Robards AW, eds.Intercellular Communication in PlantsBerlin: Springer-Verlag,197659–80.

13.

Marchant HJ. Colony formation and inversion in the green alga Eudorina elegans. Protoplasma. 1977;93:325–339.

14.

Fulton AB. Colony development in Pandorina morum. II. Colony morphogenesis and formation of the extracellular matrix. Dev Biol. 1978;64:236–251. [PubMed: 680345]

15.

Viamontes GI, Kirk DL. Cell shape changes and the mechanism of inversion in Volvox. J Cell Biol. 1977;75:719–730. [PMC free article: PMC2111588] [PubMed: 925078]

16.

Gottlieb B, Goldstein ME. Colony development in Eudorina elegans (Chlorophyta, Volvocales). J Phycol. 1977;13:358–364.

17.

Viamontes GI, Fochtmann LJ, Kirk DL. Morphogenesis in Volvox: Analysis of critical variables. Cell. 1979;17:537–550. [PubMed: 476832]

18.

Birchem R, Kochert G. Mitosis and cytokinesis in androgonidia of Volvox carteri f. weismannia. Protoplasma. 1979;100:1–12.

19.

Green KJ, Kirk DL. Cleavage patterns, cell lineages, and development of a cytoplasmic bridge system in Volvox embryos. J Cell Biol. 1981;91:743–755. [PMC free article: PMC2112818] [PubMed: 7328119]

20.

Green KJ, Viamontes GI, Kirk DL. Mechanism of formation, ultrastructure and function of the cytoplasmic bridge system during morphogenesis in Volvox. J Cell Biol. 1981;91:756–769. [PMC free article: PMC2112823] [PubMed: 7328120]

21.

Kirk DL. Volvox: Molecular-genetic origins of multicellularity and cellular differentiation. Cambridge and New York: Cambridge University Press. 1998

22.

Hallman A, Kirk DL. The developmentally regulated ECM glycoprotein ISG plays an essential role in organizing the ECM and orienting the cells of Volvox. J Cell Sci. 2000;113:4605–4617. [PubMed: 11082052]

23.

Miller SM, Schmitt R, Kirk DL. Jordan, an active Volvox transposable element similar to higher plant transposons. Plant Cell. 1993;5:1125–1138. [PMC free article: PMC160346] [PubMed: 8400878]

24.

Kirk MM, Stark K, Miller SM. et al. regA, a Volvox gene that plays a central role in germ-soma differentiation, encodes a novel regulatory protein. Development. 1999;126:639–647. [PubMed: 9895312]

25.

Miller SM, Kirk DL. glsA, a Volvox gene required for asymmetric division and germ cell specification, encodes a chaperone-like protein. Development. 126:649–658. [PubMed: 9895313]

26.

Nishii I, Ogihara S, Kirk DL. A kinesin, InvA, plays an essential role in Volvox morphogenesis. Cell. 2003;113:743–753. [PubMed: 12809605]

27.

Smith GM. A comparative study of the species of Volvox. Trans Amer Microsc Soc. 1944;63:265–310.

28.

Nozaki H. Origin and evolution of the genera Pleodorina and Volvox (Volvocales). Biologia (Bratislava). 2003;58:425–431.

29.

Hoops HJ. Somatic cell flagellar apparatuses in two species of Volvox (Chlorophyceae). J Phycol. 1984;20:20–27.

30.

McCracken MD, Starr RC. Induction and development of reproductive cells in the K-32 strains of Volvox rousseletii. Arch Protistenkd. 1970;112:262–282.

31.

McCracken MD, Barcellona WJ. Ultrastructure of sheath synthesis in Volvox rousseletii. Cytobios. 1981;32:179–187.

32.

Rich F, Pocock MA. Observations on the genus Volvox in Africa. Ann South Afr Mus. 1933;16:427–471.

33.

Mast SO. Light reactions in lower organisms. II. Volvox globator. J Comp Neurol Psychol. 1907;17:99–180.

34.

Cohn F. Die Entwickelungsgeschichte der Gattung Volvox. Beitr Biol Pfl. 1875;1:93–115.

35.

Bell G. The origin and early evolution of germ cells as illustrated by the VolvocalesIn: Halvorson HO, Monroy A, eds.The Origin and Evolution of SexNew York: Alan R Liss,1985221–256.

36.

Heinlein M, Epel BL. Macromolecular transport and signaling through plasmodesmata. Int Rev Cytol. 2004;235:93–164. [PubMed: 15219782]

37.

Mahajan-Miklos S, Cooley L. Intercellular cytoplasm transport during Drosophila oogenesis. Dev Biol. 1994;165:336–351. [PubMed: 7958404]

38.

Radford JE, Vesk M, Overall RL. Callose deposition at plasmodesmata. Protoplasma. 1998;210:30–37.

Footnotes

*

They called the alga that they found in a local pond a “globator type,” because they were not exactly certain what species it belonged to, but it clearly was a member of Smith's section Euvolvox, and for simplicity we refer to it as V. globator.