MaynardSmith_Szathmary_Ch10

MaynardSmith_Szathmary_Ch10 - Eflfifilfifilfi

Info iconThis preview shows pages 1–9. Sign up to view the full content.

View Full Document Right Arrow Icon
Background image of page 1

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Background image of page 2
Background image of page 3

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Background image of page 4
Background image of page 5

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Background image of page 6
Background image of page 7

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Background image of page 8
Background image of page 9
This is the end of the preview. Sign up to access the rest of the document.

Unformatted text preview: Eflfifilfifilfi ........................................................................................................ ._ THE EVOLUTION OF MANY'CELLED ORGANISMS An animal‘s body is composed of many millions of cells, ofmany differ- ent kinds-—rnuscle cells, nerve cells, blood cells of various kinds. and so on. Organisms of this kind have evolved independently on three occasions. giving rise to animals, piants. and fungi. The third of these groups is less complex than the other two, but a mushroom is still a fairly elaborate structure. There are in addition many kinds of simpler multicellular organisms, 1with only a few types of cells: Volvox, a hollow green sphere ofciliated cells, with germinal cells inside. is a charming example. These many independent origins ofa multiceilular mode of existence. con- trasted with the unique origin. for example, of the genetic code. of eukaryotic cells, and of meiotic sex, suggest that the step may not have been a particularly difficult one. One observation, however, points in the opposite direction. This is the apparently explosive radiation of animals at the beginning of the Cambrian period, some Sit) million years ago. suggesting that, once some crucial invention had been made. animals rapidly evolved a range of different body plans, and different ways of moving about, eating, and protecting them- selves. This raises two questions. How explosive was the Cambrian explosion? What was the crucial invention, iftliere was one? There is no doubt that, in rocks some 540 million years old, there appear for the first time abundant fossils of a great diversity of marine animals. These animals left fossils that we can find because they were large, and had shells or external skeletons that remained when their soft pans rotted away. It is clear that there were few. if any, large animals with shelis before the Cambrian, but perhaps there were small softebodied animals. We know that such animals can exist without leaving fossils—some existing kinds ofanimals have left no fossil record. This is true not only of rare and unfamiliar creatures but of some abundant and familiar ones, such as the roundworms (nematodes). which are among the world‘s most abundant animals. There is some direct evidence of animals before Cambrian times. Some 20 million years before the Cambrian explosion there are, in rocks from many 110 THE EVOLUTION OF MANY-CELLED ORGANISMS parts ofthe world. the remains ofa variety ofsoft-bodied animals, the so-called Ediacaran fauna. Although they had no hard parts to fossiliae. they left the impressions of their bodies in the mud when they died. There is debate about how these fossils should be interpreted. but they are thought to include repre- sentatives of modern phyla——certainly coelenterates [anemones and jellyfish). and probably annelids [segmented worms}. arthropods (animals with jointed limbs]. and echinoderms {starfish and sea urchins). But this takes us back ortly 20 million years before the Cambrian. Attempts to date the origin of the main animal groups by molecular means [using differences between gene sequences to measure time since divergence} suggest that many-celled animals may have originated as much as lflflfl million years ago. lfso. fossils provide information only about the last half of their history. However that may be. it remains true that something dramatic happened about 540 million years ago: many lineages of animals independently evolved large size. and hard parts. in the next section, we discuss what biological inventions were needed for multicellular life. We will be left with a puzzle. At least some of the most im- portant mechanisms are found today in single-celled eukaryotes. and even in in the physical environment making large multicellular animals possible. One suggestion is that. until about 500 million years ago, there was too little oxygen in the atmosphere and dissolved in the ocean to permit such animals to exist. The first animals could not have had a circulatory system. with blood vessels and a pumping heart: such strucmres would take time to evolve. At first. there- fore. oxygen must have reached their tissues by diffusion. a slow process. par- ticularly if the oxygen tension was low. It may be significant that most. if not all. multicellular fossils from the Precambrian era were paper-thin. Such leaf-like animals. and very small animals, could have endsted for many millions of years before they gave rise to the larger. hard-shelled animals whose remains mark the beginning of the Cambrian. It is possible that the Cambrian explosion was trig- gered by a rise in oxygen tension. aided by the emergence of one or more species of predator. whose presence made it necessary for many other animals to evolve the hard shells that constitute the major component of the fossil record. What had to he invented? Gene regulation August Weismann. who first thought clearly about genes—he called them ‘ids '—-realiaed that there were two ways in which cell differentiation could come about (Fig. It}. I }. One way. which he favoured, was that. when cells divide during development. only some genes pass to the daughter cells: thus only genes needed in the brain pass to future brain cells. only genes needed in the liver pass WHAT HAD TO BE INVENTED? 1]] (i) Figure 10.] Two lheoris of development. In {iJ. only those genes are transmitted to a particular tissue that are needed in that tissue: some gens tag. a} maybe needed in all Iissus. In {i}. all genes are transmitted to all cells, but difl‘erent genes are active [indicated by ’l in different tissues. August Weismann recognized both posibilities. but thought method ti} morelilteiynnfacnmetbodtiilistypicai. to future liver cells. and so on. The germ-line cells. destined to give rise to gametes. are the only cells that retain a complete complement of genes. as they must if they are to be the starting-point of the next generation. But Weismann saw that there was an alternative: all genes pass to all cells. but differ-ct genes are active in difl‘brent cells. To bring this about would require that influences from outside the cell. perhaps from neighbouring cells. should activate the appropriate genes. Weisrnann rejected this alternative, on the rmsonable grounds that he could not see how a sufficiently varied set of outside influences could arise. It is. however. the alternative we now believe to be correct. even though we have not hilly solved Weismann’s problem of identifying the influences that activate genes. It is perhaps worth explaining why it is that we now think Weisntann’s second. im~favoured. alternative is correct. It is not just that it looks as if the process of mitosis is designed to transmit complete sets of gena to each daughter cell. We also know that cells destined to form one structure. and which should therefore, on Weismann‘s first hypothesis, contain only the genes needed to make that structure. an. ifcircumstances change. produce a different structure. Every gardener knows that. ifone cuts oil" the tip of a plant shoot and sticks its base into sand. it will often grow roots. This only makes sense ifthe genes need- ed to make roots arepresent in shoots. Weismann was aware ofsuch facts. and ofthe difficulties theyraised forhistheorycitwasa difliculry he wasneverable to solve satisfactorily. 111 THE EVOLUTION OF MANY-CELLED ORGANISMS Transcription of the structural genes in this operon is prevented n “. "II— lil gene R Promoter Structural gene Binding of the inducer alters the regulatory protein. and it cannot bind to the promoter Regulatory . protein Transcription .1!!- _ [in gene R Promoter Structural gene Figure 10.2 The mechanism of gene regulation discovered by Francois Jacob and Jacques Mot-rod. The gene to be regulated is referred to asa ‘structural‘ gene. to it from a regulatory gene. [i] A regulatory gene {R} produces a regulatory protein. which binds to a specific promoter sequence at the start ofthe structural gene. andprevents itbeing transcribed. In iii}. an inducerbinds to the regulatory protein, and alters its shape so that it cannot bind to the protein. thus permitting transcription of the structural gene. By this mechanism. any ‘inducer‘ molecule can switch on any structural gene. Cell differentiation. then. depends on different genes being active in different cells. An understanding of how this can happen originated in the 1950s in a study by the French biologists Francois Jacob and Jacques Monod of how the bacterium Escherichia coli can acquire the ability to use the sugar lactose. The mechanism is shown in Fig. 102. The essential point is that one gene produces a protein, which recognizes and binds to a specific DNA sequence at the start of a second gene (or sometimes ofseveral linked genes l. and so is able to regulate the activity of the second gene. In the particular case of the lac operon studied by Jacob and Monod, regulation is negative: the regulatory protein switches off the second gene. unless it is rendered ineffective by binding to the ‘inducer’. lactose. In other cases. however. regulation is positive: the regulated gene is inactive unless it is switched on by the regulator gene. It turns out that such regulation is a universal feature of living cells. One WHAT HAD TO BE INVENTED? 113. feature is worth emphasizing. The ‘inducer‘. lactoushich was referred to earlier as an ‘outside influence'—binds to the regulator protein at a different site from that at which the regulator binds to the gene it is regulating A con- sequence of this. emphasized by Monod in his book. Chance and accessing is that. in principle, any chemical substance can switch on any gene. That is. the ‘meaning’ of an inducing signal is arbitrary, as the meanings of words are arbitrary. All complex communication depends on such arbitrary signals. The reader will have noticed that this example of gene regulation comes from a bacterium. In the cells of multicellular animals and plants. genes tend to have many different regulatory sequences. and are alTected by many regulatory genes. Hence the activity of a particular gene. in a particular cell. can be under both positive and negative control from different sources, and can depend on the stage of development and of the cell cycle. on the cell's tissue type. on its imme- diate neighbours. and so on. Gene regulation is complex and hierarchical Yet the basic mechanism already wrist: in prokaryotes. Cell heredity [f you take a few epithelial cells from an animal. and grow them in tissue cul- ture. they will multiply. but they will remain epithelial cells. In the same way. fibroblasts in tissue culture remain fibroblasts. and so do other kinds of diFfer- entiated cells. In other words. there is ‘cell heredity‘: like begets like. But the dif- ference between epithelial cells and fibroblasts is not caused by differences between the DNA sequence of their genes: it is caused by a difiemnce in gene activity. There is a dual inheritance system. The familiar system. responsible for hereditary transmission between generations. depends on the copying of DNA sequences during replication. The less-familiar system. responsible for cell heredity, requires the copying. during cell division. of states of gene activation. The way the second heredime system works is shown in Fig. 10.3. The activity of a gene is determined by a 'label’ attached to the gene: the best—understood labelling system is methylation. but there are others. The crucial point is that, when the cell divides and the DNA is replicated. the pattern of methylation is also copied. When gametes are produced, the labelling pattern must be restored to the initial state: in the language of computer science. a RESET button must be pressed. to restore a default state. It may come as a surprise that gene activation. determined by specific methy- lation patterns. is also found in bacteria. again. we have not discovered an invention that triggered the Cambrian explosion. Germ line and some Early in the development of all but the simplest animals. there is a division into two cell lineages: germ-line cells. which give rise to the gametes—eggs and 114 THE EVOLUTION OF MANY-CERED ORGANISMS 5, co co co 5. CG CG co 3*.=_——-_=I # 3'5! cc cc cc oc cc a: 5. co cc is If any. I Daughto'strands 3'=I [eplicafion Miriam“ at oc cc . hemimethylated sites It I «y \i . 5, co co co 5. co co cc. Unmethylated methylated 3,———' q. 3, cytosines cytosine oc cc cg or: at 6C. 1 Hemimedrylated sites Figure 10.3 Cell heredity in multicellular organisms. The figure shows methylation, the besb understood mechanism. Cytosine [C] is methylated at some G—C doublets: the particular cytosines that are methylated determine the stale ofgene activity. Cell heredity requires that the pattern iscopied when the cell divides. lnunediatelyafler DNA replication,the old strand is methylated but the new one is not. An enzyme recognizes these 'hemimelhyiated sites, and adds a methyl group to the unmethylated C. There are other mechanisms of cell heredity. depending on proteinvDNA binding. thatare lesswell understood. spenn—and hence to the next generation, and somatic cells that form the rest of the body. For Weismann. this early establishment of a germ line was neces- sary. because he thought that different somatic cells contained only the genes (ids) needed for that particular tissue: only the germ-line cells contained all the genes, as was necessary if they were to form the next generation. Now that we know that somatic cells carry a complete complement of genes. it is less obvious why such a segregation of the germ line from the soma should occur. It cannot be a necmsary feature of the development of complex many-celled organisms, because there is no segregation ofa germ line in plants: such segregation is ruled out because there is no way in which germ cells could travel from a central gonad to the flowers. where fertilization takes place. So what advantage do animals gain from the presence of a germ line? The most likely explanation lies in the nature of cell differentiation and cell heredity. We suggested in the last section that the Eormation of germ cells requires a process analogous to the pressing of the RESET button on a com- puter. All the cells must be reset to what has been called a ‘totipotent' state: their descendants must be able to differentiate into any ofthe many kinds of special- ized cells ofthe body. Ifa gamete were to arise from an epithelial cell. for example. all the gene labellingcharacteristic of an epithelial cell would have to be undone. its plants demonstrate. there is no reason in principle why this should not happen. But ii gametes are formed from undifferentiated germ-line cells, there is less need to change the labels on genes, and hence less opportunity for errors to occur. WHhT HAD TO BE. INVENTED? 115 Ways of making spatial patterns At this point, the reader could well object that all this talk of gene regulation misses the real difficulty. How does it come about that the right genes are active in the right places? How does three-dimensional form arise during development? Before we try to answer this question for animals and plants, it: will help to describe three different ways in which non‘biological forms can be made. The first mechanism is ‘template reproduction‘. An example is the production of a pattern by pressing a stamp on a piece of paper. The crucial feature is that a pre- existing form generates a copy of itself by surface-to-surface contact. A second example is the casting of a statue by pouring metal into a mould. No new form is made: a copy is made of a form that already exists. No one thinks that this is the way animal or plant forms are made. animals are not stamped out, or cast using a mould. The process is crucial in heredity. however: the replication of DNA happens by template reproduction. A second way of making a pattern is illustrated by such structures as vortices, snowflakes. or the crown of droplets formed when a spherical object is dropped on to water. In these cases, a complex and regular structure is formed by the operation of physical laws. This kind of form is often called ‘self—orgaoized’. 1t arises naturally from the properties of water—mcompressibility, viscosity, sur- face tension, and so on. One might object that the form exists only for an instant, and so cannot be an adequate model for biological form, but this is not a serious objection. There are self-organized patterns that persist: a snowflake is an example. A more serious objection is that. unlike an organism, the water splash doe not have parts, or organs, that serve to ensure its survival or repro- duction. We argued in the first chapter that adaptedness is a fundamental characteristic of life. so the objection is indeed serious. A second, related, objection is that the form is not influenced by any in- formational input: indeed, this is what is implied by the term self-organized. It follows that forms that are strictly self-organized cannot evolve by natural selec- tion, which works by altering the informational input. This is why the water splash dos not have organs ensuring its survival. There is a partial way out of this difficulty. The exact shape ofthe water splash could be altered by changing such things as the density or viscosity ofthe liquid: that is, by altering what are called the parameters of the system. Some biologists would argue that the best way to picture animal development is as a series of partly sell-organized dynamic processm like the water splash, whose parameters are controlled by genes. We thinkthis idea contains an element oftruth. For example, it is hard to think that the strips ofa zebra arise in any other way. But we also think that it leaves out a crucial aspect of development. How is it that difl'erent genes are active in dif- 116 THE EVOLUTION OF MANY-CELLED DRGANISMS ferent cells. at diFferent times. and at different places? We will return to this question. But first we describe a third way of making a non-biological pattern, in which the role of information is more explicit. A third way of generating a pattern is illustrated by computer graphics. The picture is made by an ink-jet printer attached to a computer. it stream of elec- tric impulses passes from the computer to the printer. and each black spot on the paper is formed in response to one of those impulses. Thus the pattern on the paper is generated by information that was first programmed into the com- puter, and then transmitted as a stream of electric impulses. There is a one—to- one correspondence between points on the picture and impulses in the wire. lost as no one thinks that biological form is stamped out. so no one thinks that biological form is made as a computer image is made. What is true, however. is that a protein molecule is made in away analogous to such an image. Thus there is a one-to-one correspondence between amino acids in the protein and base triplets in the gene that coded for it. Change one base and you will change one amino acid. This is not the whole story: the gene specifies the sequence of bases in the protein. but the string must then fold up to produce the three-dimen- sional form. in most cases. the suing will fold up on its own: folding is a self- organiaed dynamic process, depending on the laws of physics. which do not need to be programmed. {There is a rough analogy between the role of the laws of physics in converting a linear sequence into a three-dimensional form. and the role of the printer in convening a linear stream of impulses into a picture.) The real objection to a computer image as a model of development. however. is different. Although there is a one-to-one correspondence between triplets of bases in the gene and amino acids, there is no such correspondence between genes and parts of the body. There is not a gene responsible for the nail on your left little finger. and another for the fifteenth eyelash of your right eye. Instead. most structures are influenced by many genes. and most genes influence sever- al structures. This discussion of three examples of non-biological form may seem rather discouraging. None of them. it seems, is a satisfactory model of development. Yet. when thinking about real development. we find it helpful to have such simple models in mind. The development of organic form it is convenient to start with a simple example. In flowering plants, the flower develops from a disc of cells (Fig 10.4]. within which four concentric rings of cells differentiate. The outermost ring gives rise to the sepals, the next to petals. the next to stamens, and the central cells of the disc to carpels. Mutants are known that alter this simple pattern—for example. by converting sepals into WHAT HAD TO BE INVENTED? ll? Figure 10.4Thedevelopment ofafiowerJheflower developsfrom adisc. forrnedoffour rings ofceflaeadt dfitll'ltd to giverise todifierentstructures. petals. Some mutants of this kind will be familiar to gardeners: for example. in double paeonies the stamens have been converted to petals. Such mutants have been analysed in the simple crucifer. Arabidopsis. with the results shown in Fig. 10.5. What does this tell us or. more importantly. what does it not tell us? It shows. as we might have guessed. that the genes needed for the development of a particular structurka petal. say—are activated by particular control genes— genesaand bin theeaseofa petal Thegenesa, b, andcareat thehead ofa hier- archy of regulatory genes. But what ensures that these genes are activated in the correct regions of the floral disc. at the right time? In the case offloral develop- ment. we do not know. To answer this kind of question, we must turn to Drosophiia and the mouse, in which studies of developmental genetics have been pursued for longer. The picture becomes horrendoust complicated, but a few general principles do seem to be emerging. its August Weismann saw a century ago. we need a process whereby a group of cells. or a single cell. or ultimately a single gene, is affected by some specific outside influence. One way in which this can happen . called embryonic induc— tion. has been known for a long time. For exampie, the lens of the vertebrate eye is Formed by the difiuenfiation oftypical epithelial celis. What makes these cells different from other epithelial cells is that they come into contact with the eye cup, an outgrowth of the developing brain that will become the retina and the optic nerve. Thus a group of cells that would otherwise have become a normal 113 THE EVOLUTION OF MANY'CELLED ORGANISMS {ll (iii a‘ if“ c+ c —* Carpal be —> Stamen ix —' Stanton c —I- Carpel {iii} Figure 10.5 Normal and abnormal flower development. ti}, Development ofa flower wilh the normal genotype, u' b‘ r. Below is shown the distribution ofgene activity in the flower disc, and above are the structures that develop in response to these activities. following the mls shown on lheright. [iii Development in a flower in which the cgene is mutated and inactive: in the absence of gene c, Ihe a gene is active over the whole disc. [iiiJ Development in a flower in which the .1 gene is inactive1 and the c gene is active over the whole disc. Other mutants. including double mutants. follow the same rules. component of the skin are induced to form lens by contact with the eye cup. This has the desirable consequence that the lens forms exactly in front of the retina. A second mechanism whereby spatial pattern can arise was first suggested 3|] years ago by Lewis Wolpert. although its importance has only recently been demonstrated experimentally. First, the theon (Fig. [0.6). Suppose that some chemical substance is produced at a particular point in an embryo. The sub- stance will diffuse outwards, setting up a concentration gradient. Cells locally can respond to the concentration, and different genes can be switched on or off, WHAT HAD TO BE INVENTEDT 119 mediumF—ro low—b9 Concentration input Figure 10.6 111e ‘French Flag' model of Lewis Wolpert. a diffusible chemical substance, or ‘morphogen’, is produced at one edge ofa sheet of cells. and difl‘usion sets up a gradient of concentration. Ceflsrespond as shownto the loa] concentration, fon'ningapattemwith three dilt’erenr regions. as genes are switched on in Escherichia coli by the concentration of lactose. For example, in the egg of Drosophiio. a concentration gradient is set up of a sub— stance manufactured by cells in the ovary of the female, lying up against the eg at one pole, before it is laid. A high concentration of this substance switches on particular genes in the nuclei of the embryo, and by so doing causes that region ofthe embryo ultimately to become the head end. This example is peculiar in that the source of the gradient lies outside the embryo, in the maternal ovary. Typically, gradients arise from within the embryo itself. In principle, a single gradient could specify many embryonic regions, with different concentrations switching on different genes. In practice, however. it seems that not more than nvo. or at most three, regions are specified by a single gradient: it is interesting that Wolpert used the analogy ofthe French flag, which has only three regions. it seems that it is easier, or more reliable, for cells to respond to the presence or absence ofa substance than to many concen- tration levels. Thus only a small amount of spatial complexity is generated in a single step. Embryonic development depends on a series of steps, with genes that are switched on in one step being the source of signals in the next. The evolution of form Over the years there hasbeen remarkably little conversation between population biologists. who study inheritance and natural selection in contemporary popu- v w 9,3 1"”! r . i 120 THE EVOLUTION OF MANY-CELiED ORGANISMS lations. and palaeontologists. who study the fossil record. The former observe genetic change. the latter changes in form. If we do not know how changes in genes cause changes in form. there is not much for the two groups to say to one another. Recent discoveries in developmental genetics may bring this long separation to an end, although it must be admitted that it has not done so yet. The revolution now in progress in developmental genetics depends on new techniques in molecular biology. It is now possible, in a variety of animals and plants, to identil‘)r genes that play a part in early development, to determine their DNA base sequence and the kind of protein they code for, to find out what goes wrong if they are inactivated. to discover where and when they are first active during development. and sometimes to transfer them into distantly re- lated organisms and observe their effects. It is also possible, by looking at their sequence. to work out the evolutionary relationships between genes in different organisms. By using these techniques. much fascinating information isbeing accumulated. [t is not always easy to interpret, for reasons that can best be explained by an analogy. Imagine you wanted to find out how a motor car engine worked. but all you were allowed to do was to look at particular parts. and to destroy them and see what happened. Suppose that you removed the leads to the sparking- plugs. The engine wouldn’t Stan. The structure of the leads might suggest to you that their function was to carry an electric signal, but not much more. So what would you have learnt? Discovering how a complicated machine works by removing one part at a time and looking at it is not easy. Despite the difficulties, progress is being made. but it is perhaps too early to say what it all means. One exciting and completely unexpected finding is illustrated in Fig. 10.1 In Drosophila. there is a series of genes. known as Hox gens. characterized by the presence, at the start of each gene, of a ‘homeobox’ domain. coding For 60 amino acids These genes are active in difierent regions of the embryo. from Front to back. Fach seems to act as a master switch. activating a cascade of other genes that are needed for the development of structures appropriate to that region of the embryo. Mutations in these genes cause the appearance of the ‘wrong’ structures, or, more precisely, of structures in the wrong places. Such ‘homeotic' mutations have been known for St] yeaIs. although the genes responsible have only recently been isolated and sequenced. Clasic examples are the mutations antennapedia. which cause leg-like structures to appear on the head. in place of the antennae. and terroptera, which causes the tiny club- shaped halteres on the last segment of the thorax to be replaced by a second pair of wings. The unexpected finding is that a similar series of genes exists in the mouse and in other major groups of animals, including annelids and molluscs. The sequence of the horneobox region can be used to discover evolutionary re- ‘WHAT HAD TO BE INVENTED? 111 Figure 10.? The Hurt gene Emily. Ase-ties ofgeoes. arranged linearly alougthechromosorne of the fruit fly are active only in development. at different positions along the anteroposterior axis. and induce thedeveloprneut ofappropriatestructuros. A similar seriesof gens is active in the mouse embryo. and aiso induce appropriate structures, although the structures that deselop in mouse and fly are, of course. quite dili’erent. DNA-sequencing studieshaveshown that the most anteriorlyactinggene in the mouse is most similar tothe most anteriorlyacting gene in the fly. and so on along theseries fromlleadto taiL lationships. lt toms out that the most anteriorly acting gene in Drowplrila is more similar to the anteriorly acting gene in the mouse, and in other animals. than it is to other genes in Drasophiia, and so on down the sequence from hunt to back. What this must mean is that the common ancestor of flies. mammals. and segmented worms, and indeed of all bilaterally symmetrical animals. already possessed. some SDI] million years ago, a series of Hon: genes. acting in diHermt regions of the body from head to tail. and controlling the development of appropriate structures in those regions, and that these genes have been oun- served ever since. It has been suggested that it is the possession of these Hoar genes that is the defining characteristic of animals, in the same way that the possession of a backbone is the defining characteristic of vertebrates: the characteristic has been called the zootype. The reason why this discovery is so unexpected is that, although the H01 signalling system has been conserved, there is nothing in common between the structuru that it elicits in different groups. For example, there is nothing in fr" WDN‘V 111 THE EVOLUTION OF MANY-CELLED ORGANISMS mice, or in earthworms. corresponding to the thorax of insects, with is two pairs of wings and six legs. Of course. there is nothing mechanistically puzzling about this: one can use the same kind of switch to turn on a television set or a hair-dryer. What is puzzling is the conservation ofthe signalling system. despite changes in the structures that it elicits. and so. presumably, in the genes lower in the hierarchy. that it activates. This conservation of a signal in evolution is not unique to the Hot: genes. An equally surprising example concerns the development ofthe eye. In the mouse. there is a gene. ‘small-eye'. which. if it mutates. causes the mouse to be eyelas. Like the Hex genes. it seems to act as a master switch: when it is activated. it switches on a cascade of other genes needed in eye development. A gene with a similar DNA sequence controls eye development in Drosopitila. The startling finding is that. if the mouse gene is transferred to Dramphiln. then. wherever it is activated. it causes an eye to appear. The kind of eye that develops. of course. is a compound insect eye, not a vertebrate eye. So. when we say that the gene ‘controls’ eye development. this is rather misleading. What the gene does is to initiate a series of events at a particular site in the embryo. causing an eye to develop at that site: it does not control the kind of eye that develops. In effect. it signals 'malre an eye here’- Presumably. the common ancestor of mouse and Drosoplrila had a simple light-sensitive organ. and the position of that organ was specified by activating the gene. Now that we know about the conservation of signals. it is possible to offer an explanation. although this is beingwise after the event. A particular 1-1th gene in Drasoplriia switches on. not one other gene. but a whole ascade of genes. To alter the signal would alter many features ofthe resulting structure. Such a drastic change would be most unlikely to improve adaptation: instead. improved adaptation has been achieved by altering. one by one. the gens that rspond to an unchanging signal. The conservation ofsignals follows from the inevitability ofgradualism. ifadaptive change is to occur. A second aspect ofconservatism in morphological evolution has been recog- nized for much longer. This is the preservation of a constant body plan. despite changes. in way of life. The structural similarity between the human hand. the flipper ofa seal. and the wing ofa bat, has long been familiar: it is one of the decisive reasons for accepting the theory of evolution. Perhaps more funda- mental is the preservation of what has been called the ‘phylotype' by members of a given phylum (examples of phyla are the arthropods. chordates. annelids. echinoderms, and molluscs}. The point can best be illustrated by the chordatcs. the phylum to which we belong. During development. all chordates pass through a ‘phylotypic stage', the pharyngula. characterized by a notochord (a stiff rod. replaced later in development by the backbone]. somites {segmented blocks of tissue destined to give rise to muscles. ribs. etc}, a hollow nerve cord THE EVOLUTION OF FORM 113 dorsal to the notochord, a pharynx (the anterior part of the alimentary canal. perforated by gill slits. leading from the pharynx to the exterior). and a tail extending behind the anus. At this stage. all chordates are remarkably similar. although later they diverge: adult humans lack ootochord. gill slits, and a tail. They may also difl'er from one another earlier in development: for example, the early embryos of birds. with enormous eggs. and fish such as cod. with small non-yolky eggs. are very different. Development converges on the phylotype. and then diverges. The first point to be made about the phylotype of chordates is that it echoes the structure and way of life of our earliest ancestors. The notochord. segment- ed muscles, and post-anal tail were originally adaptations for sinusoidal swim- ming {as many fish still swim today], and the pharynx was an adaptation for filter—feeding by swallowing sea water and sieving out minute organisms as the water was expelled through the gill slits. It is interesting that some of the basic features of the body plans of other phyla are easily understood as adaptations for particular modes of locomotion; to give a second example. the liquid-filled body cavity and ring-like segments of annelids (for example, earthworms} are adaptations for burrowing. This may explain why the body plans of different phyla first evolved. but not why they have been preserved. The explanation becomes clearer ifwe look at development before and after the phylotypic stage. Before. developmental pro- cesses are global to the whole embryo and involve extensive cell migrations. In the phylotype. the main body parts are already represented by blocks of un- differentiated cells. arranged relative to one another as they will be arranged in the adult. Subsequent development is more local. with each part developing to some degree independently of the others. Evolutionary changes in development seem to have occurred in two different ways. for different reasons. Changes early in development. mainly before the phylotypic stage, involve changes in the size and yolkiness of the egg. and the presence or absence of a larval stage. Changes after the phylotypic stage lead to changes in adult structure. It may be that the conservation of the phylotype. as that of the zootype. has been forced on organisms by the need for change to be gradual ifit is to be adaptive. It is easier for mutations to cause small morpho— logical changes ifthe development of different parts is to some degree independ— ent. A genetic programme able to evolve new structures may be one that says, in effect 'set up an unchanging set of parts, with unchanging spatial relationships to one another: then, ifyou want to change anything. change one part at a time'. These ideas are vague and speculative. They are in part stimulated by research in a very different area. that of genetic algorithms. Computer scientists are con- cerned with producing optimal solutions to such problems as the design of robots. railway timetables. and power distribution networks. One way of doing "..',IL:\';/7 124 THE EVOLUTION OF MiltNY-CELLED ORGANISMS this is by evolution and natural selection. For example, a ‘genetic algorithm’ that generates a railway timetable is devised. and then allowed to evolve by mutation. recombination. and natural selection. The procedure can work very well. but it depends on devising a way of programming the task such that at least some random changes in the program lead to improvements. This is not easy. As anyone who programs in a familiar language such as BASIC or PASCAL will be aware. random changes lead almost invariably to complete breakdown. Students of genetic algorithms have thought hard about how to make their pro- grams ‘evolvable‘. The essential feature is that a small change in the program should often result in a small change in its performance: in biological language. a small change in genotype should cause a small change in phenotype {it is intriguing that computer scientists use the terms genotype and phenotype when talking about their programs}. One way of achieving this is to make the pro- gram ‘modular': that is. each part of the program should specify one. and only one, component of the total performance. It is interesting that our bodies are modular—our kidneys. livers. hearts. and legs are separate structures. perform- ing separate functions—and it is beginning to look as if our genetic programme is also to a degree modular. This. at least. is the story that the study ofembryos. and the conservation of zootype and phylotype. seems to be telling us. EEfiEIEEH ....................................................................................................... .. ANIMAL SOCIETIES Animal societies with a complex division of labour between the members are of three kinds. and have evolved by three different routes. In this chapter we discuss insect colonies—ants. bees. wasps. and tflmites—-and. more briefly, the colonies formed by some marine invertebrates. which also consist of indi- viduals specialized to perform different roles. In the next chapter we discuss human societies. Apart from the division of labour. and the economic advan- tages that follow, the three kinds of society have one other thing in common. In all cases. the different kinds of individual are genetically similar: the differences between queen and workers in a bee colony. or between farther. teacher. and shop-keeper in a human society. arise not from genetics but from the response of similar genotypes to different social circumstances. What a biologist has to explain is how a genotype could have evolved that can respond so differently to different environmental influences. The existence of non-reproductive castes, the so-called workers. in the social insects. and in some other social animals. poses a formidable problem to the theory of evolution. as Darwin already recognized. Why should worker bees give up reproduction? in what sense does this them their fitnms? A large part of the answer can be traced back to remark. by Darwin himself and later by I. B. S. Haldane. in [955: the basic idea was elaborated by William D. Hamilton in the 19605 into a general theory of animal societies. Darwin hinted that. it'the family rather than the individual was taken as the target of selection. his theory could be saved. Haldane strengthened this view by saying that he was willing to lay down his life to save two brothers. or ID cousins. His reason was that these relatives shared. on average. one-half or one-eighth of the genes poo- sessed by him. (Presumably. he said 10 rather than 8 cousins. not because he could not calculate the proportion of genes in common—he was rather good at that kind of calculation—but because he wanted to be ahead on the deal.) Why should the proportion of shared genes matter? To answer. we have to take a ‘gene’s-eye vieur‘. A gene that caused Haldane to die. but 1%] of his cousins to survive. would cause more genes identical to itself to survive than would a gene that let Haldane live and the cousins die [in fact. IDES copies of the gene. on average. would survive. compared with only one that would die). In the ...
View Full Document

This document was uploaded on 08/24/2011.

Page1 / 9

MaynardSmith_Szathmary_Ch10 - Eflfifilfifilfi

This preview shows document pages 1 - 9. Sign up to view the full document.

View Full Document Right Arrow Icon
Ask a homework question - tutors are online