sign in sign up
<<
Home , Multicellular organism

A R T I C L E S I N T E G R A T I V E B I O L O G Y 27 A R T I C L E S

The Origins of Multicellularity JOHN TYLER BONNER

There is great interest in the invention of multicellularity because it is one of the major transitions during the course of early .1 Most of the emphasis has been on why it occurred. For instance, recently Gerhart and Kirschner2 have argued that a multicellular has gained the advantage of a unicellular ancestor because it can more effectively shield itself from the vagaries of the environment by producing its own . In broader terms, this is Dawkins’3 argument that a competitively effective way of carrying the from one generation to the next is by building a complex soma that safely sees to it that the germ plasm survives.

KEY WORDS: multicellularity, differentiation, early evolution

It has been my contention for some then will see to it that spectrum. For any one population of an years that while these reasons for the the novelty is retained. I would argue organism there can equally likely be a appearance of multicellularity are con- then that the size increase came first, selection for a size decrease, should vincing, they are achieved at the most and the possible advantages that this there be an open niche for smaller forms. fundamental level by increased size and change might provide would follow. In other words, the direction of the selec- thus the key step is a transition to larger tion for size change depends solely on the size.4–7 Becoming larger makes it pos- I would argue then ecological opportunities, and there is al- sible to be isolated from the outside ways room at the top. When the was world and to be able to protect the that the size increase populated with nothing but single cells, genes for the next generation. Further- selection opportunities for multicellular- more, size increase will be accompanied came first, and the ity must have been inexhaustible. by a division of labor in the form of cell Many of the recent discussion of the differentiation, which adds to the selec- possible advantages appearance of multicellularity are con- tive advantages of the organism under fined to evolution. It is usually some circumstances. My line of reason- that this change inferred that this was a unique event ing, which will be developed further that did not spawn appreciable diversity presently, is that the first step in the might provide would until the great proliferation of type evolution of multicellularity was a size in the era. This view has increase due to an accident, e.g., a mu- follow. arisen partly because of the record, tation that prevents the daughter cells which is notably sparse until the rich from separating. If the larger cell mass findings in the Burgess Shale and simi- Another well-worn point must be has any advantages, such as ensuring lar deposits. But as someone who was stressed from the outset. One assumes the safety of the germ line by produc- raised as a cryptogamic botanist, it has the reason for the non-stop selection for ing a protected internal environment, always seemed to me that this stress on of increased size is that the animal evolution is rather anthropocen- top of the size scale is an ever-present tric. If one thinks in terms of all - open niche, and has been open during isms known to exist today, or to have John Bonner is celebrating his 50th year the entire course of organic evolution. existed in the past, then one can only be in biology at Princeton University. His Among the many lines of evidence to impressed by the diversity of the begin- major research has been in experimen- support this contention, perhaps the nings of multicellularity itself. tal studies on the development of cellu- most compelling is the simple observa- lar slime . He also has broad tion that there exists a continuum—a BRIEF DESCRIPTION OF THE interests in various aspects of evolution- complete array of sizes among organ- MAJOR ORIGINS OF ary biology, which had led to a series of isms, from the minutest to MULTICELLULARITY books from Morphogenesis in 1952 to the largest mammals and angiosperms. Sixty Years of Biology in 1996. The only place for further expansion is The danger in describing all the differ- into the unfilled niche at the top of the ent steps toward multicellularity is that the 28 I N T E G R A T I V E B I O L O G Y A R T I C L E S

walls and the division products of an asexual or a zygote remained glued together. This is well illustrated in the sea lettuce Ulva (Fig. 2) and its smaller relatives, where one can trace the transition from a simple filament to a thickened thallus. A somewhat different mode of be- coming multicellular is seen in the Vol- vocales, in which the division products are surrounded and held together by jelly and the colonies they form may be flat, for more often are hollow spheres typified by , (Fig. 3) the largest member of the group. The Chlorococcales become mul- ticellular in a radically different way. For example, in Pediastrum (Fig. 4) the products of are confined within the mother cell (in a vesicle that Figure 1. lin es th e m oth er cell). While th e daughter cells become detached and swim about using their flagella, they reader could well become overwhelmed did separate examples of aquatic origins are initially imprisoned within the by the detail. My other difficulty is that it is in the form of what is traditionally vesicle and eventually, they lose their a subject I have discussed at some length known as colonies. It is assumed that all a b ility to move, t h ey b ecom e ce- previously.4–8 Here I want to do some- higher came from green in mented into a flat plate that will burst thing different. To begin, I will adapt one which the cells had moderately rigid free from the vesicle as they grow. of the current molecular phylogenies of the major groups of organisms as pro- posed by Sogin9 (based on a single of a small subunit of rRNA), and modify his figure by including only those groups that are multicellular (Fig. 1, Drawings by Hannah Bonner). Thirteen separate in- ventions of multicellularity are indicated in the figure, but this is far below the ac- tual number because some of those eight are well known to be polyphyletic, as I will illustrate. My descriptions will be- gin with and plants and move clockwise around the figure. The descriptions themselves will be brief, and to make it easier to fix them in one’s mind, some of the more important forays into multicellularity will be ac- companied by drawings. The reader may want to skim over this section just suffi- ciently to appreciate the variety of sepa- rate ways single cells have evolved into multicellular organisms.

Green Algae and Green Plants

The green algae began their multicellu- Figure 2. larity in water. They provide some splen- A R T I C L E S I N T E G R A T I V E B I O L O G Y 29

The question why there is such a mentous in , and they are to vary- variety of shapes in these algae was ad- ing degrees syncytial or multinucleate, dressed many years ago by Baker,10 who especially in the growth phases of their pointed out that photosynthetic algae cycle. One very common aspect of did not need to develop an elaborate their multicellularity is that when they feeding mechanism to catch particu- go into their reproductive, or fruiting late food; all they needed was to be phase (Fig. 5), all the nuclei and cyto- able to catch the sun, making it pos- plasm suddenly flow through the hy- sible to invent a great variety of dif- phae to central collection points and ferent shapes, which is exactly what rapidly produce spore-bearing fruiting bodies. These can be quite small, as in they have done. the case of simple molds, or they can be very large mushrooms. In the case of animals, we know little Figure 3. about the transition stages and are left with a huge gap between the and the ancestral single cells. Either all Hydrodictyon is closely related to the known , living and fossil, Pediastrum, but differs in that it produces, came from one multicellular ancestor, or by much the same development, a huge possibly there was more than one ances- . Finally, I should mention the co- tor, sponges having had a separate origin. enocytic and multinucleate green algae such as Caulerpa, a large marine form Fungi that is attached to the ocean floor by a holdfast. Inside there are no cross-walls, Fungi are a heterogeneous group and but merely a large vacuole surrounded the possibility that they are invented by streaming cytoplasm containing vast multicellularity more than once is a rea- Figure 5. numbers of nuclei. sonable hypothesis. All fungi are fila-

Brown Algae and Diatoms The are largely marine and are notable for having some forms, such as Macrocystis, which are over 100 ft in length. It is assumed that their method of initially becoming multicellular was much the same as I have already de- scribed for the green alga Ulva; that is, they started as cells with rigid walls that failed to separate upon division. Diatoms, which are related to the brown algae, would seem to be quint- essential unicellular forms; each cell is encased in a hard silica shell. There are, in fact, two minor but interesting excep- tions within the group. There are a few in which the dividing cells re- main attached at one end to form sessile colonies. The other exception is particu- larly odd: all the motile cells secrete a tube that surrounds them and expands as they multiply. This tube, which branches, is anchored to the ocean floor. The se- creted house may be a centimeter or Figure 4. more in height, and the separate cells actively move about inside it (Fig. 6). 30 I N T E G R A T I V E B I O L O G Y A R T I C L E S

Figure 8.

Red Algae Although have many multicel- lular forms that are shaped like some of the green and brown algae, their cyto- logical and biochemical characteristics are quite distinct and set them apart. Figure 6. Nevertheless, it is assumed that they achieved their multicellularity in the same fashion previously described for the other algae. Because they differ in so many details of their cell structure, they must have made the step to multi- cellularity independently. The cell structure of ciliates is highly specialized and different from all other Cellular Slime Molds organisms. Ciliates are unique in having These organisms are characterized by an a separation of the germ line micro- asexual life cycle in which they feed as nucleus, and a huge that separate amoebae. When their food sup- controls the morphogenetic events of ply is consumed they aggregate into the complex cortex. It could be argued collections of cells that form small, mul- that this is another way of becoming ticellular fruiting bodies bearing resis- larger—one that avoids multicellularity. tant (Fig. 9). However, there are a number of genu- ine multicellular, or colonial, forms which exude a supporting adhesive thread at one end of the dividing cells, ultimately building a sessile colony of individual but connected cells. In Zoothamnium (Fig. 7) the cells are also linked by a muscle thread so that if one cells is touched the whole colony will contract to avoid danger. One particularly curious form of multicellularity is found in Sorogena (Fig. 8) a that in the soil. When its food has been depleted, the separate cells aggregate to form a small fruiting Figure 7. Figure 9. body that sticks up in the air. A R T I C L E S I N T E G R A T I V E B I O L O G Y 31

The cellular slime molds consists of cies of form groups of two major groups, the Acrasids and the compacted cells that seem to adhere Dictyolstelids. They clearly have separate closely to one another after division. It origins for their differ in their cell struc- has been pointed to me by Prof. Karl ture and are quite far removed in a mo- Stetter that this organism is an obligate lecular phylogeny.11,12 anaerobe and that these compact colo- nies might serve as a mechanism to keep the internal cells protected from a sudden Myxomycetes influx of oxygen in the environment. True slime molds are quite different than the cellular slime molds. They have a Eubacteria: sexual cycle in which the zygote begins as an uninucleate , but as it When these motile rod-shaped bacteria feeds and becomes larger only the nu- grow and divide the daughter cells re- Figure 12. clei divide and it develops into a large, main close to one another, forming a multinucleate mass of naked proto- wandering swarm that steadily in- plasm—a . When the con- creases in size. When the conditions are are a number of species, with branching ditions favor fruiting, the plasmodium right these masses of rods come to- filaments. Although they are primarily breaks up into a single aggregate, and gether in central collection points to aquatic, they are extraordinarily hardy more often many small ones, each of fruit—to form cysts or microspores. In and can withstand considerable expo- which forms a fruiting body bearing some species, such as Chondromyces sure to air. They do form resistant spores haploid spores that give rise to the ga- (Fig. 11), the cysts are lifted up into the and since they are photosynthetic they metes of the next generation (Fig. 10). air on a stalk. also have specially differentiated cells (heterocysts) to fix nitrogen from the air (Fig. 13).

Sexuality vs. Asexuality Let me add a comment here concern- ing all of these different explorations into multicellularity. Unlike so many ad- vances in evolution, does not seem to have been a prerequisite for the in- vention of multicellularity. Clearly sexu- ality exists among unicellular organisms; in other words, sex undoubtedly ante- Figure 10. Figure 11. dated multicellularity. However, there is no obvious correlation between invent- ing multicellularity and sexuality. The Foraminifera and Radiolaria Actinomycetes primitive multicellular organisms may These are amoebae with beautiful shells These soil bacterial form small, branch- produce either asexual spores or ga- made with calcium or silica. They mostly ing thread-like filaments. Some of those metes. Among the fungi and the green float free in the ocean—a few are sessile. filaments will reach up into the and pro- algae are species in which there is an As they grow they become multinucle- duce spores. Often the spores bud off a alternation of sexual an d asexual ate and in the foraminiferans they se- linear series, although in some species cycles, both of which will produce crete additional chambers to their shells there is a spherical mass of spores at the spores or some form of unicellular with size increase. tip of a filament. They are a large and propagules. diverse group; perhaps the best known Archaebacteria is Streptomyces (Fig. 12), which produces MECHANICAL WAYS OF the antibiotic streptomycin. BECOMING MULTICELLULAR Archaebacteria are a newly identified group of prokaryotes that are quite If we now look at these various experi- distinct in many of their biochemical ments in inventing multicellularity it is characteristics from eubacteria. These The cells of these eubacteria are large— intriguing to ask if some general state- ancient organisms are notable in that as large as many eukaryotic cells. They ments about their different mecha- they can exist in many extreme environ- are mostly multicellular, the usual form nisms can b e m a d e. Th ere is o n e ments. There are reports that some spe- being linear filaments; however, there particularly important point that I 32 I N T E G R A T I V E B I O L O G Y A R T I C L E S

their cells as building blocks once they fail to separate. Animal cells do not have a and they stick together because of ad- hesion molecules on their surface mem- branes. That they are covered only by a membrane is no doubt related to the fact that originally the cells were amoe- bae and they had to retain a pliable membrane so that they could engulf bacteria and other food particles. Con- sistent with Baker’s10 point mentioned earlier, the next big mechanical step must have been devising a feeding mechanism for a group of cells, but any clue as to how this was achieved initially has been long lost. Diatoms and ciliates have highly specialized outside coverings that present another kind of problem. In the case of diatoms, it is the rigid silica shell, Figure 13. while in ciliates, it is the elaborate cor- tex and its cilia and other complex struc- tures; in both cases, an external armor have made earlier.8 One can identify a All the aquatic raises a challenge to . Yet a sharp distinction between those or- few species of both groups have found ganisms that became multicellular organisms began an identical solution. By exuding an ad- under water and those that did so on hesive stalk at one end of the elaborate land. All the aquatic organisms began their multicellularity cells they can divide without interference their multicellularity by the products from the adhesive, and the daughter cells of cell division failing to separate, by the products of will be attached to one another at one while most terrestrial end—by their roots, so to speak—and in involve some form of motile aggrega- cell division failing to this way form a branching colony. tion of cells or nuclei in a multinucleate Then there is a genuinely curious . There are some apparent ex- separate, while most solution, also previously described, of ceptions such as the actinomycetes the diatom that surrounds itself with a and a few cyanobacteria, e.g., some terrestrial branched secreted tube as it multiplies. species of Stigonema, which live in air It is as though it were creating a minia- in moist environments. However, it is microorganisms ture biosphere. most likely that the ancestors of these particular forms were aquatic. The ter- involve some form of Terrestrial Origins restrial origins are indicated in italics and the aquatic origins in regular type motile aggregation For microorganisms it is difficult to draw (Fig. 1). a sharp line between aquatic and terres- of cells or nuclei in a trial, because the latter need water too and will always exist in a thin layer of Aquatic Origins multinucleate water covering particles of soil or humus. In the case of the aquatic origins there In other words, all terrestrial microbes is great variety and this variation seems syncytium. are to some degree aquatic. This is es- to be correlated largely with the type of pecially true of their feeding phase, for cell, and especially the type of cell sur- almost none of the multicellular terres- face. For those organisms with cell walls and ultimately the branching and thick- trial microorganisms are photosynthetic one of the most prevalent methods is to ening of those filaments to form solid and they all require a liquid film. If they have the cells fail to separate after divi- tissues and produce large plants. The are particle feeders they need this to sion. The hard polysaccharides of the cell same mechanism in its filamentous form move about and engulf their food. If walls found in green, brown, and red al- is found in the cyanobacteria and some they are saprophytes they must get the gae have permitted the rise of filaments other eubacteria. All of these forms use food itself in liquid form. A R T I C L E S I N T E G R A T I V E B I O L O G Y 33

These terrestrial organisms send a we saw in fungi and myxomycetes. The large indeed, as in a giant puffball that may spore-bearing body that pushes up multinucleate mass in myxomycetes, weigh pounds). through the water interface into the air which can be very large, is covered by only for dispersal. This is certainly true for the a thin membrane and therefore requires fungi with their enormous array of aerial very moist conditions on the forest floor WHEN MULTICELLULARITY fruiting bodies (although the water or on dead logs. This plasmodium is the OCCURRED IN EARTH molds have similar sporangia that form feeding stage, which is quite separate in HISTORY and liberate motile zoospores below the time from the fruiting or dispersal stage. There is a general point about the ori- surface of water). It is equally the case The very same separation of the gins of multicellularity that needs em- for both kinds of cellular slime molds as feeding and fruiting phases is found phasis. It is obvious that the moment in well as the true slime molds, or myxo- among the great multitude of species of evolutionary history for each of the dif- mycetes. Aerial spore-bearing bodies are fungi. However, there is one major dif- ferent organisms described here un- also characteristic of terrestrial multicel- ference. The syncytium is always con- doubtedly occurred not only as separate lular prokaryotes: some myxobacteria fined in a chitinous tube, the hypha. As events but at different times. There is product fruiting bodies with prominent good evidence that multicellularity in stalks while others merely form mounds, In the case of cyanobacteria was invented at a very and the actinomycetes have chains of early stage, a good 3.5 billion years spores sticking up into the air. aggregating single ago,16 yet the first multicellular animals The mechanisms of aggregation, was probably a much more recent event. which is characteristic of most terrestrial cells, we have a Unfortunately, we have no way of know- microorganisms, takes two forms: it in- remarkable bit of ing the sequence in time of the origins volves either the gathering of separate of the various groups of organisms uninucleate cells or the gathering of convergence, for shown in Figure 1, but they must span nuclei and cytoplasm in a multinucleate an eon. Indeed there is nothing to rule syncytium. aggregation occurs out the possibility that at this very mo- In the case of aggregating single ment multicellularity is in the process of cells, we have a remarkable bit of con- independently in being invented by some single-cell form vergence, for aggregation occurs inde- somewhere on our earth. pendently in eubacteria, in two distinct eubacteria, in two kinds of cellular slime molds, and in cili- ates; in other words, it has arisen inde- distinct kinds of WHAT MIGHT BE THE pendently at least four times. In addition, SELECTION PRESSURES there are numerous variations in the cellular slime molds, THAT PRODUCE mechanics of fruiting body formation and in ciliates; in MULTICELLULARITY? within some of the groups. In the cellu- Aquatic Origins lar slime molds, for example, the spores other words, it has can be pushed up as the cells squirm The fact that multicellularity arose inde- past one another, rising into the air; by arisen independently pendently so many times is the primary creating a stalk of dead cells made by basis for believing that there has been a adding live cells at the tip which become at least four times. significant selection for it in the ancient trapped and vacuolate; by rising on a unicellular world. Yet it is difficult to slender, cell-free of cellulose also guess that the first advantages might secreted by the cells at their apical end; these hyphae branch and spread into have been. In the case of aquatic origins, and by secreting a non-cellular cone and the mycelium, they will invade the soil it is easy to imagine that the failure of using osmotic pressure to push the or rotten wood, whatever might be its the daughter cells to separate might be spores up to its narrow ends like an substrate, sopping up nutrients, and the result of a simple . It is the erupting volcano. This latter mechanism continuously expanding and enlarging. next step that is harder to picture: what is interesting because it is somewhat When the right conditions arise, all advantages would clusters of cells have similar to the method used by the totally growth ceases, and the protoplasm over single cells? Perhaps initially they unrelated ciliate, Sorogena; again a strik- surges back through the hyphae to a cen- had neither advantage nor disadvan- ing case of convergence (see Fig. 8). tral collection point (see Fig. 5). It may be tage and survived by drift until some In those instances where there is a a modest surge and produce a simple further mutational change endowed multinucleate syncytium the aggrega- spore-bearing mass held up in the air by a them with a skill that was not possible tion stage is quite different; it involves single hypha, as in Mucor, a bread , or for their single-cell relatives. I will pur- the directed movement of a great mass it may involve the aggregation of a large sue this possibility in the next section; of protoplasm toward central locations amount of protoplasm in a vast mycelium here I am only concerned with the ini- that are the incipient fruiting bodies as to form a mushroom (which can be very tial step. 34 I N T E G R A T I V E B I O L O G Y A R T I C L E S

It might be that the mutation that ing bodies suggests that there has been fruiting body. We do not know enough allowed the cells to adhere to one an- and still is an enormously strong selec- about dispersal mechanisms in the soil, other also allowed them to stick to the tion pressure for the dispersal of spores, and can only ask whether this primitive substratum. Under some circumstances, cysts, and even seeds in higher plants. clustering of cysts might somehow en- where the cells in an ideal location for At first glance it would seem that, as be- hance dispersal. growth are likely to be swept away by fore, this development of a fruiting body In the case of syncytial forms, the currents, remaining fixed to one spot must have been something that oc- multicellularity is more closely associated might be selectively advantageous. The curred well after the appearance of mul- with feeding. In both the myxomycetes (a same advantages would apply to the ticellularity, but let us examine the ) and the myxobacteria (a daughter cells, thereby giving rise to a possibilities a bit more carefully. ), it is clear that by an increase multicellular sessile colony. Such an oc- In the case of cellular slime molds, in size of the feeding mass they can feed currence might be the origin of colonial there is a unicellular relative, Protostelium more effectively. They produce extracellu- stalked ciliates and diatoms. This could (Fig. 14), that makes its own stalk so that it lar that digest large particulate even apply to the diatom that is encased rises up into the air to form a single spore. food which they then absorb directly. In in a tube—instead of having the adhesive Clearly, if a fruiting body could be myxobacteria, Dworkin15 has called this glue the cells to the substratum, they had made with numerous cells it might be wolf-pack feeding. So in both of these a different set of that allowed even more effective in dispersal. For a cases we could guess that multicellularity them to build a cocoon of stiff material soil amoeba, where feeing must be done arose as an advantage in feeding, and the around them, inside of which they could as single cells by , the ag- formation of fruiting bodies was second- divide and grow. In these instances, be- gregation of cells is required to achieve arily derived because of the advantage of coming multicellular is the inevitable con- multicellularity. The question of how this effective dispersal. The very same argu- sequence of the advantage of remaining ments would apply to the fungi. in one spot. Becoming I have said very little about foramin- CAPITALIZING ON iferans and radiolarians. They are mostly multicellular opens MULTICELLULARITY free in the ocean, or pelagic, and as they grow they become multinucleate. But the gateway for all Becoming multicellular opens the gate- what is the advantage in their doing so? way for all sorts of remarkable innova- One argument has been put forth by sorts of remarkable tions that would be impossible for single Bell13 for Volvox that might apply here: cells. By being larger and by being made the size increase prevents filter feeders innovations that up of numerous cells, organisms can from being able to eat them. The prob- have a division of labor and also respond lem for this ingenious argument is that would be impossible to their environment in new and sensi- we are looking for the origins, and the tive ways, all that have led “too big to eat”’ idea requires that for single cells. to their success. There are many possible there are already large multicellular examples to illustrate this point; here I predators about. Perhaps initially pe- will give two. arose is more difficult, but a species of lagic multicellularity had no advan- the distantly related soil amoeba Hart- tage—they just grew. manella forms resistant cysts in clus- Division of Labor in In the case of primitive flagellated ters.14 There is an aggregation of the Cyanobacteria or ciliated cells, it is conceivable that size amoebae before encystment, yet no increase is an advantage, because the The first example comes from the cyano- larger the organism the faster it will bacteria, known from the fossil record to swim. Again it is easy to see that in an be a very ancient group. They have two aquatic world where there is a multitude biochemical functions that are immis- of different-sized organisms, being large cible, which has led to a division of la- and fast helps to catch prey (or to escape), bor. Cyanobacteria are photosynthetic but again this does not help us explain the organisms and therefore produce oxy- first step. There lurks the possibility that gen in the presence of sunlight, and at initially there was no advantage and it was the same time they must fix the free ni- only later that it was retained because of trogen from their immediate environ- positive selection pressure. ment, an absolute necessity to make their proteins and all the other compo- nent molecules in their body that con- Terrestrial Origins tain nitrogen. Since nitrogen fixation can The incredibly large number of different Figure 14. only take place in the total absence of organisms existing today that have fruit- oxygen, the two processes are essential A R T I C L E S I N T E G R A T I V E B I O L O G Y 35 for their existence cannot occur simul- mass, pick up the sticky spores and carry aker and Poff19 showed that this was taneously in the same place. Many spe- them to some virgin patch of bacterial only true if the slugs are migrating in a cies of cyanobacteria have solved the food. Now I want to show that these temperature range above that in which problem by doing their in simple bags of aggregated amoebae do they had been raised; if the gradient was the daytime and their nitrogen fixation at much more than just stick their spores in a colder range they would go toward night, but here I am more interested in a up into the air to be tagged by passing the colder side, i.e., they are negatively second and more sophisticated method, beasts: they go to quite extraordinary thermotactic. Whitaker and Poff19 gave where some of the cells in the filament lengths to see that the spores end up in a neat and convincing explanation to become specialized for nitrogen fixation. the optimal place for dispersal. I will not this reversal of orientation: in daytime These so-called heterocysts are give all the experimental evidence here the sun would make the air and the sur- clear cells, lacking chlorophyll, and have for the basis of this assertion; I have re- face of the soil generally warm so the thick walls that prevent the oxygen from viewed it elsewhere.8,18 Here I will only migrating slugs would crawl toward the neighboring cells from getting into their give the results and describe the skills surface, but at night the heat gradient is inner machinery. By this cellular division of the aggregated cell mass. reversed and the soil will be warmer of labor cyanobacteria can take in en- than the cool night air; yet since in a ergy from the sun and use it to make These simple bags of cooler environment they are negatively nitrogen compounds at the same time thermotactic they will still go upward. By they are pulling in nitrogen from their aggregated amoebae this intriguing mechanism they orient surroundings. The nitrogen products are correctly toward the surface both night passed along to the photosynthesizing do much more than and day. cells through the pores at the end walls It is also known that cellular slime of the heterocysts; all the cells therefore just stick their spores molds orient by exuding a gas which not benefit from their work, and they in turn only orients slugs, but more importantly receive the nutrients they need from the up into the air to be positions the rising fruiting body. This neighboring green cells. gas, which we now know is ammonia, Over and above these remarkable tagged by passing speeds up the cells, so if two fruiting facts, it is known that in some species bodies arise close to one another they the heterocysts are perfectly spaced beasts: they go to will lean away from each other. This is along a filament of cells. Wilcox et al.17 because the ammonia that they give off have shown that they give off an in- quite extraordinary is more concentrated between them, hibitor that diffuses along the filament and the cells on the inner side of each which prevents any cell for the limit of lengths to see that will move faster and cause the rising its effective from turning into cells masses to move away from each a heterocyst. So despite the fact that the spores end up in other. One can also show that they not they are simple prokaryotes they have only avoid other fruiting bodies but for differentiation into two cell types (in fact the optimal place for the same reason they will move away three, because some cells become resis- from a wall—in fact, it is this mechanism tant spores for surviving hard times), dispersal. that makes them rise at right angles and that differentiation is organized into from the substratum. If they are in a a regularly spaced multicellular pattern. small cavity in the soil they will, by this None of these achievements could have It would appear that most of their gas orientation, position themselves in arisen in a unicellular cyanobacterium. abilities center around getting the the dead centre of the cavity, and all spores from the deeper moist feeding these sensitive orientations contribute to putting the spore in the ideal pace for Behavior in Cellular area nearer to the surface of the soil where apparently there is a better dispersal. These are remarkable feats for Slime Molds chance of catching a ride. The migrating a bag of amoebae. In the cellular slime molds, spore dis- cell masses are extremely sensitive to From all this we see that during the persal appears to be of great importance light and will go toward light of surpris- long course of evolution the numerous for their reproductive success. So far I ingly low intensities. Since daylight or experiments in becoming multicellular have said only that sticking up into the moonlight will always be from above, were really just small steps compared to air is the key—in this way the spores will this is one powerful way to orient them the wonders that followed. more effectively spread. There is reason upward in the soil. to suspect that near the soil surface and There is more. They are also highly CONCLUSIONS in the humus where these organisms, sensitive to heat gradients and will ori- passing animals such as mites, worms of ent in gradients as small as 0.5°C/cm. We What we see in this examination of the various kinds, and other motile inverte- first thought orientation only occurred large array of experiments in multicel- brates that brush against the spore toward warmer temperatures, but Whit- lularity is that in early evolution becom- 36 I N T E G R A T I V E B I O L O G Y A R T I C L E S ing larger took on many forms; in fact, Adhesion plays a role in aggregation too. 3 Dawkins R (1976) “The Selfish Gene.” Oxford: there are no doubt others we do not Finally, there is an especially fascinating Oxford University Press. 4 Bonner JT (1958) “The Evolution of Develop- even know about which have gone ex- question: in the first signs of cell differ- ment.” Cambridge: Cambridge University Press. tinct. As I have argued, the most reason- entiation there is both a mechanism to 5 Bonner JT (1965) “Size and Cycle.” Princeton, NJ: able guess is that originally they arose alter the fate of an individual cell and a Princeton University Press. by chance mutation, and subsequently mechanism to place the differentiated 6 Bonner JT (1974) “On Development.” Cambridge, were selected because of some advan- cells in a regulated, controlled pattern. MA: Harvard University Press. 7 Bonner JT (1988) “The Evolution of Complexity.” tage that they might accidentally have What are the similarities and differences Princeton, NJ: Princeton University Press. occurred. This initial success was often in the molecular mechanisms of all the 8 Bonner JT (1993) “Life Cycles.” Princeton, NJ: greatly improved upon, as we saw for independent inventions of these re- Princeton University Press. cyanobacteria and cellular slime molds, markable phenomena? I think the study 9 Sogin ML (1991) Early evolution and the origin but the raw materials for natural selec- of the origins of multicellularity has a of the . Curr Opin Genet Dev 1:457–463. 10 Baker JR (1948) The status of the . Na- tion had been laid down by cells acci- bright future. ture 161:548–551, 587–589. dently clumping together. 11 Olive LS (1975) “The Mycetozoans.” New York: These days we have become be- ACKNOWLEDGMENTS Academic Press. guiled with diversity: how animals, such 12 Raper KB (1984) “The Dictyostelids.” Princeton, as insects, and plants, such as angio- I particularly thank Laura Katz, who was NJ: Princeton University Press. , have produced so incredibly of enormous help in the initial prepara- 13 Bell G (1985) The origin and early evolution of many species. In the origins of multicel- tion of this paper. Many thanks as well germ cells as illustrated in the Volvocales. In Halverson HO, Monroy A (eds): “The Origin and Evo- lularity we see a most primitive example to Hannah Bonner for her fine drawings. lution of Sex.” New York: Alan R. Liss, pp 221–256. of diversification. In some ways, it is al- Finally, I thank Peter Kareiva, Jonathan 14 Ray DL, Hayes RE (1954) Hartmanella astronyxis: most an ideal case because we can make Weiner, and an anonymous reviewer for A new species of free-living ameba. J Morphol an argument for its basis: size increase is very helpful suggestions. I received in- 95:159–188. the common cause of all the small suc- formation on terrestrial cyanobacteria 15 Dworkin M (1972) The myxobacteria: New di- cesses that I have described in this essay. through the kindness of S. Golubic and rections in studies of procaryotic development. It is this very cellular diversification L. Margulis. The content of this paper was Crit Rev Microbiol 2:435–452. 16 Schoph JW (1993) Microfossils of the early that will be so interesting to examine in initially presented at the Forum on the apex chert: New evidence of the antiq- our future inquiries. What are the bio- Origin of Life in Göteborg, Sweden, in uity of life. Science 260:640–646. chemical differences and similarities April 1997, by the kind invitation of Prof. 17 Wilcox M, Mitchison G, Smith RJ (1973) Pattern among the adhesives that are used in Clas Blomberg and the Swedish Science formation in the blue green alga, Anabaena. I. Ba- the various aquatic forms that invented Research Council. sic mechanisms. J Cell Sci 12:707–723. multicellularity? Is there a genetic con- 18 Bonner JT (1994) The migration stage of nection between the separate ones, or : Behavior without muscles or nerves. REFERENCES FEMS Microbiol Lett 120:1–8. are some or all of them unique? The 19 Whitaker BD, Poff KL (1980) Thermal adapta- same questions can be asked about 1 Maynard Smith J, Szathmary E (1995) “Major Transitions in Evolution.” New York: W.H. Freeman tion of thermosensing and negative thermotaxis in terrestrial forms: what are Spektrum. in Dictyostelium. Exp Cell Res 128:87–93. the molecular connections, if any, be- 2 Gerhart J, Kirschner M (1997) “Cells, , tween the different chemoattractants? and Evolution.” Oxford: Blackwell Scientific. © 1998 Wiley-Liss, Inc.