Click for next page ( 141


The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 140
5 Development The Molecular Mechanisms by Which Organisms Develop Can Now Be Investigated Since the time of Aristotle, a major preoccupation of biologists has been the description of how an organism develops from an embryo to its adult form. By the beginning of this century, the elaboration of the cell theory, the discovery of the details of fertilization, and the development of improved histological tech- niques had led to an accurate description of the anatomical and cellular details of development in many kinds of organisms. As they learned more about the elaborate processes of gastrulation, neurulation, and pattern formation, however, biologists yearned for mechanistic explanations. Emerging theories emphasized the importance of the structure of the egg, the lineage of cell divisions, the accurate timing of these divisions, the importance of cell-cell interactions, and the role of inductions. More recently, with the advent of new techniques for dealing with the genetics and molecular biology of developing systems, this essentially anatomical description has been extended to a detailed analysis of selective gene expression and to the discovery of genes that regulate developmental decisions. Despite this progress, a molecular explanation for the processes of develop- ment remains an elusive goal. In many branches of biology, phenomenological accounts of life processes have been replaced by detailed chemical descriptions; many complex processes, such as DNA replication and virus assembly, have been reconstituted from purified components in vitro. Comparable results have been difficult to achieve in developmental biology, despite their fundamental impor- tance for advances in the field. Giant strides may at last be possible in developmental biology, however, because advances in cell and molecular biology, as well as in the precise study of 140

OCR for page 140
DEVELOPMENT 141 developmental systems themselves, have greatly improved understanding of the properties of eukaryotic cells. In the field of cell biology, cellular components such as microtubules and actin are now reasonably well understood. Differential cell adhesion can now be described in terms of specific molecules, receptors, and known elements of the extracellular matrix. Communication between cells, hitherto mysterious, can be explained in terms of such components as soluble growth factors, second messengers, cell receptors, and cell junctions. Molecular biology has likewise contributed much to the improved prospects for an understanding of development. For example, cell differentiation can now be understood primarily as differential gene expression, coupled with the modifi- cation and turnover of macromolecules. In addition to providing sensitive probes for following developmental events-essential for biochemical work with indi- vidual embryos molecular biology has begun to provide us with an understand- ing of how gene expression is controlled during development. Our understanding of the links that connect cell morphology, the extracellular environment, and gene expression is still incomplete, but we have already learned much about how metazoan developmental systems function. Armed with these new techniques, many scientists have begun to address the classical problems of developmental biology with renewed vigor. A number of different systems have been investigated, each yielding important contributions to our understanding of the overall problems involved. In Drosophila, for example, the combination of genetics with molecular biology has led to the discovery of important regulatory genes. Other organisms have been studied because of the regularity of their cleavage process. At the same time, the classical objects of developmental studies, such as sea urchins and frogs, have continued to reveal important facts about the physiology of fertilization, the regulation of gene expression, morphogenetic movements, induction, and cell-cycle regulation. Our expanding knowledge has made it clear that the basic cell biological processes of development are common to all organisms, so that the combination of these results has proved especially fruitful. This chapter is designed to provide a selective tour through the sequence of fundamental developmental mechanisms, emphasizing the interplay among them. Differentiation of two highly specialized cells, the egg and the sperm, contains the instructions for the earliest steps of development. The tightly controlled interac- tion between these gametes, which is called fertilization, breaks the developmen- tal arrest that is characteristic of germ cells and initiates other controls of cell growth and cell division. As the single fertilized cell divides, the daughter cells begin to differentiate along separate pathways by expressing different subsets of genes from their identical genomes. The unfolding of the genetic program of each cell is directed by both internal and external cues. Internal cues include the poorly understood "determinants," which are inhomogeneously distributed in the oocyte and distributed unequally to certain cells as the embryonic cells divide. External cues are derived from interactions with other cells and with extracellular matrices

OCR for page 140
142 OPPORTUNITIES IN BIOLOGY laid down by other cells. The external information is enriched by very specific cell movements that allow different kinds of cell-cell interactions during different developmental periods. The emphasis of this chapter will shift among different organisms and differ- ent techniques, but is designed to explicate our understanding of the major developmental mechanisms in cell biological and molecular terms. The prospects for advance in this field are extraordinary, with its major questions approachable today in ways that they could not have been conceived of even a few years ago. These new investigations will not only begin to answer the intellectual questions that have preoccupied scientists and philosophers since at least the time of Aristotle, but they hold the promise of considerable practical impact on our understanding of the entire field of biology, especially on medicine and agricul- ture. An understanding of the way an organism develops is of fundamental importance to comprehending and utilizing the properties of that organism. DEVELOPMENT BEGINS WITH GAMETOGENESIS Understanding the Differentiation of Germ Cells Will Give a Key to the Initial Steps of Development In one sense, development begins with the productive encounter of the sperm and egg at the time of fertilization. However, fertilization actually has its basis much earlier, during the long and complicated process of the growth and develop- ment of that sperm and egg. Sperm and eggs develop from a specific group of cells in the embryo called germ cells. In some organisms, the precursors of the germ cells can be identified and followed throughout the course of an embryo's development. Eliminating the precursors from a fertilized egg renders He result- ing animal sterile. It is lilcely that specific molecules in the egg, the so-called determinants, are responsible in such cases for the development of the germ cells. The molecular nature of these determinants and the mechanisms of their action are unknown; however, transplantation of cytoplasm containing these determi- nants can induce the ectopic development of germ cells. In other kinds of organisms, the presence of specific determinants has not been detected; in these cases, it is unclear whether determinants, comparable to those in the animals in which they are known, are actually absent or simply not detectable by available methods. In mammals, germ cells can be identified at a certain point in develop- ment; the origin of these cells is uncertain, however, and the mechanisms by which they appear are unknown. In most animals, the germ cells differentiate along two different pathways, one of which results in the production of a large, immobile gamete called the egg, while the other leads to the production of a small, mobile gamete called the sperm. Mature gametes contain half the amount of DNA (half the number of chromo- somes) present in somatic cells; their union restores the diploid number of chro

OCR for page 140
DEVELOPMENT 143 mosomes. In most animal species, eggs and sperm usually reside in separate female and male individuals, respectively. In each sex, the differentiation of germ cells into mature gametes is a process of key importance since it represents the basis for understanding the initiation of development; such understanding is likewise of great practical importance. The Egg Contains Both Nutritive Materials and Positional Information Needed for the Early Stages of Development The process of oogenesis, or egg production, results in cells that contain sufficient stored material to support at least the first stages of development. In addition to nutritive and structural materials, the egg also contains information necessary for directing subsequent development. This information is produced and appropriately distributed according to instructions included in the egg genome and is also influenced by the contributions of other maternal cells surrounding the egg. Information is stored in different kinds of molecules, mainly proteins and RNAs, some of which are nonrandomly distributed in the egg cytoplasm. Other molecules, less well known, probably also contribute to the information pool. In addition, the plasma membrane of the egg contains information in molecules unequally distributed on the surface. Furthermore, the development and matura- tion of the egg is controlled by numerous outside factors (mostly hormones), which are essential for normal oogenesis, although they may not contribute directly to the complexity of the egg structure. By synthesizing the appropriate receptors, egg cells can regulate, to a degree, their responsiveness to these outside factors. Oogenesis is a unique process, combining cytoplasmic diversification with significant growth. In nonmammalian species, fully grown oocytes contain all of the raw material needed to support embryogenesis. Even in mammals, in which embryonic growth is largely supported by materials from the mother, the egg is several hundredfold larger than somatic cells. Although the egg contains a tremendous amount of information, this information alone is not sufficiently complex to direct every detail of development. Instead, subsequent development is a series of interacting processes that call upon the genome and the existing maternal materials to generate further complexity. It is the proper understanding of this series of reactions, whereby the crude information of the egg is trans- formed into the detailed information of the organism, that constitutes the main goal of developmental biology. The enormous growth of the oocyte occurs while the cell is arrested in meiotic prophase. In order to continue development, the cell must be released from its state of arrest to complete the meiotic divisions. In mammals this process is under the control of gonadotropin hormones; this control is likely mediated by the follicle cells surrounding the oocyte. How they do this remains one of the most important questions in reproductive biology. The recent development of in

OCR for page 140
144 OPPORTUNITIES IN BIOLOGY vitro techniques that enable at least partial reproduction of oocyte growth and meiotic maturation should contribute significantly to the understanding of the biochemical basis of this process. In addition, these techniques will significantly enhance our ability to control the reproductive capacity of mammalian species; their clinical, agricultural, and ecological applications can be readily visualized. Some progress has been made in identifying the genes that are active and necessary for oogenesis. For example, in Drosophila, many of these genes direct the first stages of the program of morphogenesis. Sperm-Specific Changes in Chromatin Structure Seem to Control Gene Expression The spermatozoon contributes half of the genetic material to the developing individual and provides the necessary stimulus for the activation of development in the egg. During spermatogenesis, the chromatin undergoes unique changes as the protamines replace histones; this process is reversed after fertilization. Since the changes in chromatin structure are probably associated with changes in transcriptional activity, one can speculate that gene expression in the sperm genome is specifically controlled. Unique methylation changes occur during spermatogenesis in many species, and these changes might cause specific infor- mation storage in the sperm DNA. The morphology of the adult sperm is species-specif~c and varied. Sperm morphology, controlled in part by the sperm genome and in part by supporting cells, is probably another example of complex interactions between the process of development and particular sets of genes. Although it is unclear whether sperm provide any nongenetic information crucial for development, the fact that the eggs of many animal species can undergo parthenogenetic development-develop- ment into individuals of normal appearance in the absence of fertilization- argues against this possibility. The absence of parthenogenesis in mammals, however, might suggest that the contributions of the sperm and egg are different and that both are essential for normal development. Recent results indicate that during mammalian oogenesis and spermatogenesis, the genome of the gamete undergoes different imprinting that modifies its activities. The mechanism and extent of this imprinting are unknown and should be studied since they represent an important mode of gene control. The Interactions Between Sperm and Egg Are Tightly Regulated at Several Levels During the course of evolution, mechanisms have originated that ensure the specificity of the sperm-egg interaction and prevent the fusion of more than one pronucleus from each parent in the formation of the zygote. After being released from the male genital tract, sperm undergo a final maturation step (capacitation),

OCR for page 140
DEVELOPMENT ~ : DO THE EGG AND SP~ERM;CON I RIBUTE EQUIVALENT GENETIC : i: : INFORMATION ~3 THE ANIMAL EMBRYO? : ~ In ~m~am~mals~the~egg:~and::sperm:~:do ndt~contribute: equivalent Sets o f genet~:information eve:n~:~;byond :the; obV~us~difere~nces contributed by the: : . ::~ sex: chromosomes. :~M~rom:anipu~lat:ion techniques have~mad~e it possible to: ::: r~place~:the male pronucleus~he sorb nucleus Abhors fus~n~in:a:~mouse : z ygote: w~h~t;he female~p~ronucleus from another zygote: and ~'ice~versa~. ~ In this way ~ has been: possi:ble~:to Study the: ~pre:lmplantation devebpment~ of :: constructs embryos containing Am: s~ts::of:fem:al~e~:~(biparental gynogenomes) or two sets~of male ~ipa:rerital~androgenomes) genomes. Airing :: : implanted gynogen~etic embryos fail to develop the extraembryon~ic compo- nents~he trophoblast and yo;lk Sac whereas androgendt~ embryos fail to develop the embryo proper. Both classes of embryos eventually abort:: ~ : indicating that male and::female genom~es Oust Present and~that each performs a different but essential role in development. Gent experiments have l~al~zed~ fubct~n~al differences between Male and female genomes to 145 :: :: :: several chromosomes or subchromosomal regions. It should~now be: pos- ~ sible to identify those genie sets~that are expressed differently in the chrome- : unman: ~ ~ntr~buted bv the male ::and female. Knowing what make these: identical ge~natic~ elements function differently depending on their parental derivation is a maffer~of great theoretical and practical importance. ~ : : which renders them competent to fertilize the egg. Sperm-egg interactions are probably regulated by specific recognition signals, and several molecules in- volved in this process have now been characterized. Both capacitation and sperm-egg interaction are important targets for contraceptive intervention and must therefore be studied further. Great progress has been made over the past decade in our understanding of the physiology of fertilization, largely as a result of improvements in techniques to measure small changes in ion and lipid concentrations. The process of fertiliza- tion represents a case in which two cell types that have spent some time in a quiescent storage state must rapidly resume a high level of metabolic activity and fuse with each other. This change in state has been termed activation. In both eggs and sperm, activation is accompanied by a dramatic change in intracellular ion concentration, particularly that of Ca2+ and H+ (more commonly referred to as pHi). These changes must be important steps in activation because altering either PHi or [Ca2+li tends to stimulate the parthenogenetic development of invertebrate eggs. The Ca2+ ionophore A23187 is a general activator of eggs, apparently

OCR for page 140
/ 146 OPPORTUNITIES IN BIOLOGY because of its ability to elevate [Ca2+li. In contrast, ammonia and other weak bases, which elevate pHi, activate some but not all of the events of early develop- ment, including DNA synthesis, chromosome condensation, and increased rates of protein synthesis in the sea urchin egg. Changes in PHi and [Ca2+]i are potent effecters of metabolism and develop- ment in numerous species, in different cell types, and at various developmental stages. Such changes accompany normal fertilization in many eggs, but the mechanisms by which the changes are achieved are poorly known. In the case of [Ca2+]i, a specific Me of lipid in the plasma membrane of the eggs of both invertebrates and vertebrates (phosphatidylinositol-bisphosphate or PIP2) rapidly cleaves after activation, releasing inositol trisphosphate (IP3) into the cytoplasm, while diacylglycerol (DAG) remains in the plasma membrane. Both of these molecules are important second messengers since IP3 triggers Ca2. release from intracellular stores and DAG activates an enzyme called protein kinase C, which turns on other important cellular proteins by phosphorylating them. This chain of events leads to an increase in [Ca2+]i. Perhaps the most interesting aspect of this story is that this same process of inositol lipid cleavage occurs in a wide variety of other cell types that respond to the binding of a variety of external agents, including hormones, growth factors, neurotransmitters, and chemotactic mole- cules at their plasma membrane. Thus studies of fertilization may help to explain signal transduction in other systems and serves as a model system for understand- ing the role of second messengers in other kinds of systems. CELL DIVISION, GROWTH, AND DEVELOPMENTAL TIMING Cell Division and Cell Growth Must Be Exquisitely Regz~atedfor Normal Development Regulating cell division and growth together is a requirement for homeosta- sis. If growth outpaces division, the cell will become larger and larger. If division is faster than growth, the cell will become smaller and smaller. All eukaryotic cells exhibit a chromosome cycle and a cytoplasmic cycle, which are well corre- lated with one another. The chromosome cycle includes a period of DNA replication (S phase), during which each gene is duplicated, and a period of mitosis (M phase), during which the genes are segregated to the daughter cells. In the cytoplasm, growth is more or less continuous and the bulk content overall is duplicated; only the centrosomes undergo a discrete duplication (usually in S) and a segregation (in M). In most cells, the S phase is separated from the M phase by periods during which neither replication nor segregation occurs. The period between M and S (when a eukaryotic cell is typically diploid) is called G1, whereas the period between S and M (when a cell is typically tetraploid), is called G2.

OCR for page 140
DEVELOPMENT 147 Many important problems in biology, medicine, and agriculture concern how this simple cell cycle functions. In the embryo and in renewable tissues, such as bone marrow, cells are continually entering this cycle. Most differentiated cells, however, are not dividing and are resting in G1. Some cells, such as mature nerve cells of the central nervous system, will rest in G1 for the life of the organism. Cells in some tissues are only contingently arrested in G1. For example, cells such as the fibroblasts in the skin or parenchymal cells in the liver are capable of entering the cell cycle if the organ is injured. Under such circumstances, these normally quiescent cells initiate DNA replication and divide until sufficient cell replacement has occurred. The inability of some cells, such as nerve and muscle cells, to reenter the mitotic cycle limits the capacity of their tissues to recover from injury. At the other extreme are tumor cells, which reenter the mitotic cycle too easily and do not respond to the normal signals that arrest division. The specific events of the cell cycle are also critical points for the life of the cell. Errors in DNA replication cause mutation, and errors in meiosis or mitosis cause chromosome abnormalities. In embryonic development, growth is precisely regulated. For an animal to develop, specific cells must divide at a specific time and cease to divide at a specific time. Some cells are even programmed to die. Recent studies in the development of the embryo of a nematode, a member of a simple group of worms, give a dramatic example of the regularity of the growth process. Each individual is almost identical to the next. The cells divide on schedule and with a reproduc- ible orientation. Certain cells keep dividing while others cease dividing, and some die on cue. In human development, although the regularity is less extreme (presumably other factors ensure the successful production of the embryo since our pride would never allow us to admit that a nematode is put together more exquisitely than a human), it is likely that the same principles underlie the regulation of cell division in both organisms. Not until recently has our understanding of cell division progressed apprecia- bly beyond the descriptive stage, and we are just beginning to understand the underlying biochemical mechanisms. Simpler systems such as yeast are being studied to enhance our knowledge of the principles involved, but embryonic systems have also proved useful for this purpose. The egg and early embryo of many animals are essentially nongrowing systems, dividing their cytoplasm into increasingly smaller cells. In such systems, only the pace of the cell cycle, and not the accumulation of a certain minimum cell mass, determines when division occurs. In such embryonic systems, intracellular factors that regulate the cell cycle have been described for the first time. Improved methods for protein separation, immunological characterization, and cloning have recently increased our understanding of the extracellular growth factors that may cause quiescent cells to begin to divide. We now know a considerable amount about what these factors are, but still have little knowledge of how they act after reaching the plasma membranes of their target cells.

OCR for page 140
148 OPPORTUNITIES IN BIOLOGY Many fundamental questions about the way in which cell growth is regulated during the course of embryology are under active investigation. For example, we would like to know, What are the factors that time cell division and other events? How are they segregated equally to daughter cells? What causes specialized cells such as neuroblasts to proliferate, and what causes them to cease proliferating? What prevents cell death in one daughter cell and yet causes its sibling to die? How is cell-cycle timing related to differentiation? Are other developmental events tied to the cell-cycle timer in the way our morning alarm is tied to a clock? These old questions are currently benefiting from new investigations in which results from different organisms are being elegantly tied together to produce new results of generality. As we shall see, the story of the current investigations into the cell cycle carries us from tumor viruses to nematodes and from yeast genetics to frog embryology. It is a search for new molecules and a continuing search for the ways old molecules function. Yeast Provide a Good Model System for Studying Molecular Mechanisms of Cell Division Yeast are unicellular fungi, simple in some respects, but with many of the same features that characterize mammalian and all other eukaryotic cells. Like all euk~uyotes, yeast cells have a nucleus, chromosomes, mitotic activity, and a cell cycle divisible into G1, S. G2, and M. The growth of yeast cells, like that of other eukaryotic cells, can be arrested in G1 if they are deprived of nutrients; similarly, growth can be arrested in the same stage by factors that specifically inhibit division, such as the mating pheromone. Yeast cells can easily be manipulated genetically, and many mutants that cause cells to arrest at specific points in the cell cycle have been identified. In recent years the study of the cell cycle in baker's yeast, which divide by budding, and fission yeast, which divide like nor- mal mammalian or plant cells, has yielded important genes that are involved in control of the cell cycle in all higher organisms. Starting with the striking obser- vation that a human gene can replace a yeast regulatory gene controlling cell division there has been a steady stream of reports showing homologies between human and yeast structure and function. Recently these studies in yeast have been combined with those in frog, sea urchin, and human to reveal some of the basic workings of the control mechanisms for the cell cycle. The yeast cell may be of particular utility in demonstrating feedback control of the cell cycle that occurs when there is damage to DNA or when mitosis is inhibited. The future will see many opportunities to exploit the genetic advantages in both budding and fission yeast with the biochemical advantages of other systems. It will be of particular interest to combine studies of cell signaling through the mating pheromones with the cell-cycle arrest that ensues. This may turn out to be a good analogy with the many examples of control of the mammalian cell cycle by growth factors or other

OCR for page 140
DEVELOPMENT 149 extracellular regulators. However, to exploit the potential fully will require the development of in vitro systems for yeast. Study of Embryonic Cell Division Is Leading to a Better Understanding of the Cell Cycle in All Cells A toad embryo does not pause to grow. Instead it divides at an incredibly rapid rate, with a cell cycle 25 times as fast as normal somatic cells in culture. The Xenopus egg, for example, does not pause between M and S or between S and M during its first 12 divisions; yet before fertilization the cell cycle has been shut down completely. The large amount of cytoplasm that these eggs possess makes them ideal subjects in which to distinguish the contribution of the cytoplasm from that of the nucleus. Experiments on these eggs demonstrate that the regulating machinery of division is in the cytoplasm; the nucleus is merely a responding element. Several years ago, a factor was identified from frog eggs that induced meiosis when injected into frog oocytes. This factor, called maturation-promoting factor, was subsequently shown to be present at mitosis of all eukaryotic cells. In vitro experiments have been carried out in which nuclei have been reconstituted from DNA and soluble components and induced to break down and undergo mitosis by the addition of maturation-promoting factor. The complex steps of nuclear assembly and disassembly may soon be reducible to pathways similar to those which have been demonstrated for virus assembly. In the case of the cell cycle, however, the assembly process is carefully regulated. We know that the final reaction regulating the assembly and disassembly of the nuclear envelope is phosphorylation of a specific protein. From a different direction, other embryonic studies have provided another piece to the solution of the puzzle of the cell cycle. Although embryonic cells are endowed with a generous supply of all known structural proteins needed for cell division, they must still synthesize proteins in order to pass through the cell cycle. This has suggested to some that protein synthesis serves a regulatory function, a view that was reinforced when the amounts of specific proteins were found to oscillate during the embryonic cell cycle in clam and sea urchin eggs. These proteins, named cycling, induce cell-cycle transitions in frog eggs, which tie them to the mitotic factors described above. In the past two years, work on maturation- promoting factor from frog eggs, cycling in sea urchins, and regulatory genes in yeast have come together in a spectacular way. Maturation-promoting factor has been purified and found to contain as its principal component the homologue of the yeast regulatory gene. Cyclin activates maturation-promoting factor from a stockpile of inactive material. The basic workings of the cell cycle have been reconstituted in extracts in which the intimate biochemical relationships between cyclin and the yeast gene can be studied. All of these basic systems have been

OCR for page 140
150 OPPORTUNITIES IN BIOLOGY immediately applicable to all organisms. The use of advanced molecular biologi- cal techniques, like the polymerase chain reaction, makes it possible in a matter of weeks to identify genes from one organism to another across hundreds of millions of years of evolutionary history. It is thus no longer necessary to find the optimum system for study; one can use eggs for biochemical analysis, yeast for genetics, and Drosophila for developmental studies. In the future, this approach will become more prevalent. Progress on control of the cell cycle will be accelerated by assuming that everything basically works the same. Differences will emerge, but the power of assuming the basic conservation of fundamental biological processes like cell division will be more and more obvious. Growth Factors and Their Receptors Play Many Roles in Shaping the Organism Extracellular factors that stimulate the entry into a proliferating stage of the cell cycle have been much better characterized than intracellular factors. Some, like epidermal growth factor (EGF), have been purified for many years, and their receptors in the cell membrane have been well characterized. Recently, the connections between oncogenes and growth factors and oncogenes and growth- factor receptors have become clearer. For example, several oncogenes and growth-factor receptors are tyrosine kineses. Since most oncogenes have a cellular homologue, their study gives us new approaches to normal cell-cycle regulation. At present the number of growth factors and receptors is unknown. Some are clearly tissue specific; others, such as EGF and insulinlike growth fac- tors, are widely distributed. Our knowledge of growth-factor receptors and soluble factors has two impor- tant gaps. First, even when the activity of a receptor has been identified (for example, Hat it is a tyrosine kinase), we do not know the substrates for this activity; even in those cases in which we can identify some of the substrates we do not know how they induce cell proliferation. The second major gap is our knowledge of the role of the growth factors in normal development. These factors modify other properties of tlie cell in addition to proliferation. Some, for example tumor growth factor, control the production of extracellular matrix. Others, such as fibroblast growth factor, induce the proliferation of blood vessels. The classi- cal embryology literature is replete with descriptions of factors, often obtained from heterologous sources, that will induce the formation of early tissue types, such as mesoderm or neural tube. It seems likely that some of these factors will ultimately prove to be known growth factors or proteins encoded by proto- oncogenes. Although the well-studied growth factors have been soluble proteins, em- bryological studies have provided evidence that the extracellular matrix can be an important inducer of differentiation events and can control cell proliferation. Recently, several extracellular matrix proteins have been shown to have se- quences related to EGF. The potential importance of interactions of these kinds in

OCR for page 140
164 OPPORTUNITIES IN BIOLOGY Many questions remain for the future. But with the advent of monoclonal antibody and recombinant DNA technologies, the events of cell movement, cell adhesion, and cell recognition during development can finally be studied at the level of molecular mechanisms. With the discovery of a number of major cell and substrate adhesion molecules and their receptors, and with the recent progress in uncovering additional adhesion and recognition systems, we can expect great advances over the next decade in our understanding of what mechanisms control these basic events of morphogenesis and how these events help control the devel- opment of tissues and organs throughout the body. POSITIONAL INFORMATION Species Differ in the Patterns in Which the Cell Types Are Arranged, Not in Their Cell Types What distinguishes one group of organisms from another, and indeed one part of an individual organism's body from another, is the way in which cell Apes are arranged with respect to one another. The mechanisms that operate during development to ensure the correct spatial arrangement of cells, tissues, and organs are included in the term pattern formation. Although no unified view of pattern formation yet exists, an understanding of the behavior of embryos and their constituent cells is recognized as necessary to deduce probable mechanisms. Probable mechanisms can then be tested experimentally; if they survive such tests, they can be used to guide the formulation of questions about the molecular nature of patterning mechanisms. Although the problem of pattern formation lies at the heart of not only developmental biology but also evolutionary biology, it is a late-bloomer com- pared with other problems in development. The current status of the field has been likened to that of genetics at about the time of Mendel. However, consider- able progress has been made in recent years, and advances in understanding pattern formation promise to occupy center stage in developmental biology in the next decade. Diverse model systems have proven advantageous for investigating the vari- ous aspects of pattern formation. There are four major episodes of patterning events. 1. The placement of the cytoplasmic constituents important to subsequent development during oogenesis, most likely in a spatially organized pattern. In response to fertilization, this early pattern is extensively and precisely reorganized . . in many species. 2. The establishment of the main axes of the body in the multicellular embryo, which has emerged as a consequence of repeated divisions of the egg. In accomplishing this feat, cells meet different fates along both the anterior-posterior and the dorsal-ventral axes of the body. During this process, cells initiate exten

OCR for page 140
DEVELOPMENT 165 sive movements to bring previously separated regions of the body into proximity, and patterning information is transferred (induced) between newly opposed sheets of cells. 3. The development of appendages, such as legs and wings, at particular positions along the main body axis. 4. The development of patterned structures at the body surface, such as scales, hairs, and feathers. Does Cell Lineage Completely Determine the Fate of a Cell? Perhaps the simplest of the views of animal development is one that consid- ers the egg to have highly detailed information sufficient to specify the features of the adult. This view has found support in such well-studied embryos as those of leeches and nematodes, in which cell lineage is normally invariant during devel- opment. Further experimental analyses, however, have cast doubt on the validity of this simple view of things. For example, in Xenopus embryos, the pattern of cell lineage is precise enough to enable a detailed fate map of the major body parts to be constructed from early cleavage stages. Nevertheless, experimentally pro- duced variations in this cleavage pattem, while altering the"standard" fate map, have no consequence for the emergence of the final form. In fish embryos, early cleavages seem to yield reproducible patterns of cell lineage, but these bear no relation to the final pattern of the body; during early gastrulation, prior to the establishment of the main body axis, cells from different lineages migrate indi- vidually and mix randomly. Even in nematodes and leeches, in which cell lineage under undisturbed conditions is invariant, examples are accumulating that suggest that cell lineage and the determination of cell fate are not obligatorily coupled. These studies suggest that the environment outside a cell, whether this is other cells or molecules, is important in the emergence of pattern even in situ- ations in which cell lineage predicts cell fate. But only when it is possible to follow the fates of cells isolated from early cleavage stages, either alone or after transplantation to a new site, can the extent to which the environment affects early development be assessed. However, evidence from a variety of embryos suggests the presence in eggs of determinants that become segregated into certain lineages. The best-studied example of such a determinant is the one that specifies the germ cell lineage in a variety of animals. The emerging view is that localized determinants exist, but that specify key positions within the embryo are fewer than would be required to specify the entire body pattern. Patterning of the cells that lie between the specialized key positions (for example, between the extreme ends of the body) involves interactions, either short- or long-range, between cells. We anticipate that in the next few years, the molecular identity of at least some determinants will be uncovered and will lead in turn to an understanding of how cytoplasmic determinants interact with the genes to determine cell fate. In addition, the existence of precise information about normal cell lineage, made possible by a battery of new techniques for cell

OCR for page 140
166 OPPORTUNITIES IN BIOLOGY marking, will be of great assistance in unraveling the patterning events that do not depend on the inheritance of determinants. How Can the Environment of the Developing Cell Play a Role in Deters ng Its Fate? Development proceeds with increased spatial complexity in the embryo. Most recent studies have concluded that, although there may be some prelocalized information such as RNAs in the egg, the spatial complexity of the egg is fundamentally simple, being confined to localization of information in the ante- rior and posterior poles of eggs such as frog and flies. As cells divide, the embryo becomes more complex and most of this complexity arises through cell-cell inter- actions. Evidences that cells influence and transform the fate of their neighbors goes back to early experiments on the induction of the axial organization of frogs. Such inductive interactions have been found in all organisms. The search for the inducers was a history of frustration until recently. It was long known that dead or heterologous tissues had potent capacity to induce new tissue types. Only recently have known growth factors been tested in the appro- priate assays and been shown to be extremely effective in eliciting induction. With sensitive molecular biological techniques it has been possible to examine embryonic tissues, where it has been found that embryos in the earliest stages of development contain both the mRNAs for growth factors and the molecules themselves. In the frog the mRNA for a relative of a known growth factor in- volved in wound healing has been found localized in the egg in a region where the earliest inductive signals are generated. Most of the signaling molecules act locally, which is consistent with the behavior of embryos in classic transplantation experiments. In some cases there is good evidence that molecules exist, called mo~hogens, that act over a long range and provide positional information for tissue organization. The best candidate thus far for a morphogen in vertebrate systems is retinoic acid. This lipid-soluble compound, derived from vitamin A, can have dramatic effects on cells. For example, low concentrations of retinoic acid cause tera- tocarcinoma cells in vitro to differentiate into heart muscle cells, whereas high concentrations favor the differentiation of neurons. Most importantly, in a num- ber of different systems, exogenously applied retinoic acid and its analogues seem to affect patterning dramatically, and research is currently directed at determining whether or not retinoic acid acts as a morphogen during normal development. Even if retinoic acid is not an in viva morphogen, it will be very useful as a probe because once we understand how it alters patterns, we will know more about the patterning process itself. Local Cell-Cell Interactions May Play a Large Role in Pattern Formation Patterns may emerge as a result of local cell-cell interactions by a process of intercalation rather than as a result of long-range signaling by morphogens. Inter

OCR for page 140
DEVELOPMENT 167 calation occurs when cells that are normally not adjacent come into contact, either as a result of reanangements during development or wound healing or as a result of grafting. This contact stimulates cell division, which continues within the system until all the intervening structures are replaced by the proper pattern. In insects, the epithelial cells of the imaginal disk carry out intercalation. In am- phibians, connective tissue fibroblasts play this role, as is most clearly seen in the regeneration of lost limbs. After removal of part of the appendage, wound healing brings normally nonadjacent cells into contact, producing a discontinuity in the normally smooth gradation of positional values, which stimulates cell division and intercalation to reduce the discontinuity. The studies on amphibians raise the intriguing possibility that mammals may one day be stimulated to regenerate their limbs if a way can be- found to reactivate the developmental programs used for forming limbs in the embryos. Most of the experimental evidence on pattern formation comes from regener- ating systems; it is not yet clear to what extent intercalation may establish and regulate the primary body pattern of animals. Future research is needed to specifically address the issue of cell-cell interactions and intercalations during de- velopment of the early embryo. Research on pattern formation is aimed at understanding how cells, as the units of development, interact with one another and their environment in produc- ing the characteristic patterns of organisms and their parts. Answering this question will require not only understanding which genes are active at what times, but also appreciating what activities the cells are engaged in and how gene activity relates to this. For example, unequivocally identifying a morphogen in the near future would not by itself explain pattern formation, just as knowledge about insulin and its structure have not explained its mode of action. Complete understanding will come from knowledge of the molecules that act as signals in conjunction with knowledge about the responses of cells to these signals. DEVELOPMENT IS FOR ADULT ANIMALS TOO The processes of development do not cease with the hatching of an egg or birth of an animal. Metamorphosis in insects and amphibians, limb regeneration, and even the attainment of sexual maturity by mammals during adolescence, are illustrations of development as a process that continues throughout life; develop- ment could even be considered to include the controlled phenomena of death. How Is Cell plumber Controlled in Different Tissues of the Adult? A number of materials that control the activity of partially differentiated stem cells have been discovered. For example, the recently cloned colony-stimulating factors stimulate the production of macrophages and granulocytes, and erythro- poietin stimulates the production of red blood cells. Such systems cannot simply be maintained by inductive signals, however, since each type of cell has a characteristic lifetime and concentration in the body. Homeostatic mechanisms

OCR for page 140
168 OPPORTUNITIES IN BIOLOGY must measure such numbers and control the totals of each type of cell present. Neutrophils, for example, have a half-life of about 5 hours. Our bodies contain constant numbers of these cells, averaging about 4 x 10~ per individual. The numbers rise in response to bacterial infection or shock and eventually return to their steady-state levels upon recovery. Something in the individual must be monitoring and controlling these numbers, yet at present we understand little about the process. Other cells of the hematopoietic system have different lifetimes and inductive signals; they are presumably monitored by other systems. Likewise, nearly every part of the body that is subject to renewal in adult life must have some kind of homeostatic monitoring system, from the cell lining of the intestine, which turns over rapidly, to the liver. In adult rats, liver cells (called hepatocytes) turn over relatively slowly: they have half-lives of about 7 days. If 90 percent of the liver is removed surgically, however, hepatocyte division is rapidly induced, and the organ is restored to its original size within a few days. The regenerating liver does not significantly overshoot or undershoot its size goal when reconstituting itself; this indicates the existence of accurate mechanisms for measuring the proper size of an adult rat liver and inhibiting hepatocyte proliferation as this size is reached. Again, little is known about how this mechanism might work. It might monitor blood levels of metabolites handled or produced by the liver. Or, it might measure the size of the liver by means of specialized junctions between cells, called gap junctions. Gap junctions are small channels in cell membranes that connect neighboring cells in venous tissues, including the liver. These junctions allow the free diffusion between cells of small molecules; theoretically, the concentrations of such molecules within the liver cell could serve as a measure of the total size of the organ. What Genetic and Physiological Mechanisms Determine the Life-Span of Cells and Organisms? A chapter on growth and development would be incomplete without men- tioning aging and death. Accidents aside, biological processes control these events, both at the level of the whole organism and at the level of the individual cell. Not only do different species live for different lengths of time, but even within a given species, such as the laboratory mouse, different strains have different life expectancies. Obviously there must be genetic influences on life- span. Some mouse strains are susceptible to diabetes, autoimmune disease, neuromuscular problems, or particular types of cancer; these diseases contribute to shortened lifetimes in these strains. Such differences, however, cannot account for the fact that mice consistently live for shorter times than some related species of rodents, such as rats.

OCR for page 140
DEVELOPMENT 169 Even within a given organism different cells have vastly different life-spans. Certain nerve cells exist for as long as the individual itself. Other kinds of cells, such as neutrophils and intestinal cells in vertebrates, turn over rapidly. Even individual cells of a particular type may survive for different lengths of time depending on other events. For example, mammalian T lymphocytes are formed in the thymus. Within the thymus, 95 percent of such cells die rapidly, within two or three days of their formation, unless they are selected by the thymus because of a particular specificity of their receptor for antigen. Successfully selected cells are released by the thymes and migrate to other parts of the body, where they become part of the large pool of T lymphocytes in the animal, responsible for fighting off infections. Even there, these cells have a relatively short half-life- less than a week-unless they encounter an antigen to which their receptors can bind. If this happens, the T lymphocyte divides and produces various hormonelike factors (lymphokines), which stimulate the B cells (antibody-producing cells) to divide and help rid the animal of the invading antigen. Once the invader is destroyed some of Be progeny of the once-dividing T lymphocyte become "memory cells"; they stop dividing and producing lymphokines, but they survive in He animal more or less indefinitely, with a life-span approximately that of the individual itself. By this means the immune system builds up a pool of long-lived T lymphocytes, which are useful in fighting off the types of infections that its host will encounter during life. A single encounter with an antigen, and burst of cell division, changes the life expectancy of the human T cell, without any further cell division, from less than a week to more than 10 years. SPECIAL PROBLEMS IN PLANT DEVELOPMENT Plants and Animals Share Many Structures and Developmental Mechanisms, But There Are Some Major Differences The fast and most obvious difference between plant and animal development is that plant development is usually repetitive in nature and indefinitely long. A tip of a maple twig will put out pairs of leaves all season long and then again the following year. Roots are less periodic in structure, but they too grow and branch indefinitely. The plant body is basically the accumulation of the products of its past developmental activity. Many of the cells remain alive. By design, however, many do not, resulting in the accumulation of wood and bark. Occupying ever more volume and intercepting ever more light is obviously a key strategy of high adaptive significance for plants. This strategy is generally implemented by having the tip regions elongate in length indefinitely, a process called primary growth. Subterminal regions stop elongating and, in perennials, increase in girth at their periphery, while producing dead but functional wood cells toward the interior of the stem or root. The well-known rings in cross

OCR for page 140
170 OPPORTUNITIES IN BIOLOGY sections of woody stems that grow in areas with a seasonal climate are the result of this secondary growth. It is not too far off the mark to characterize plant development as continued, or repeated, embryology. A second special feature of plants is that the cells do not move relative to one another. Even the male gamete lacks flagella in most groups of contemporary plants and is borne to the vicinity of the egg within a pollen tube. In plant embryology, there is no phenomenon comparable to the sudden contact of a group of cells with a new cellular environment, as occurs in animal gastrulation. None- theless, plants exhibit diverse cell types and complex morphogenesis. A third major difference between development in plants and animals is that in plants the germline is not distinct Cells in many different parts of the plant as in flowers on many branches-may undergo meiosis. The products of meiosis are not gametes, as they are in animals, but rather haploid spores, which divide mitoti- cally, forming a haploid phase in the life cycle called the gametophyte. In many ferns and bryophytes, the gametophytes are green, photosynthetic, and free- living, whereas in seed plants flowering plants and gymnosperms-they are highly reduced, enclosed within and completely dependent nutritionally on the sporophytes on which they are borne. Describing the Sequential Details of Development and Experimentally Modifying This Sequence Are the Major Approaches to Studying Plant Development Important data on plant development have been obtained by using clonal analysis. Here x-rays induce visible heritable changes in individual cells, tagging them and all their progeny through further development. These clones are ideal for cell lineage studies, and they can reveal how many cells are involved in the formation of a given plant organ, such as a leaf. The intriguing result is that the number is never 1. The initiation of an organ is a "group donation" of 10 to 20 cells. Tissue character can be independent of cell lineage: epidermal cells normally divide as a coherent surface sheet; however, when an epidermal cell oc- casionally divides and contributes a cell to the interior of the leaf, the cell differentiates as an interior cell. Other studies have shown how the cytoskeleton changes in relation to cell differentiation and to the initiation of organs. Through in situ hybridization of nucleic acids it is possible to find out which cells make specific transcripts and when. Spatially and temporally defined patterns of transcription have been found, among other places, in the interaction between pollen and stigma surfaces that precedes pollen-tube growth and fertili- zation in the flowering plants. This phenomenon is discussed in more detail in Chapter 1 1. Developmental mutations offer the prospect of revealing key steps in the developmental chain, where the presence or absence of a single kind of protein determines a major change in the course of development. Mutations of this sort

OCR for page 140
DevELoPME~r 171 include those which convert the normally complex, bilateral form of a snapdragon flower to a radially symmetrical one, like that of a morning glory. Another set of well-defined mutations converts a compound pea leaf into a set of tendrils or makes the leaf into a set of round stalks. Thus single genes can profoundly affect organ character. Other mutations put the right organ in the wrong place. Certain mutants in Arabidopsis have extra sets of stamens, have petals in place of stamens, or exhibit other deviations from the normal condition. These are homeotic mutants, with well-known equivalents in Drosophila and other invertebrates. Other mutants disrupt the timing of developmental events. Light effects on plants, beyond photosynthetic effects, have been well known since the time of Darwin. Such effects, along with the hormone activity that underlies them, are discussed in Chapter 11. Plant scientists have succeeded in identifying many key control points along the causal chain from genome to a full- grown plant. In general, the nature of the agent with an effect for example, mutation, light absorption, hormone structure is well understood; the nature of the responding system, on which the agents act, is not. Some of the central research opportunities in the field of plant development center on supplying this missing information. Plant Cell Growth The Plant Cell Stands as a Key Intermediate Unit in the Sequence from Gene to Phenotype On the one hand, the genome produces a cell with a repertoire of physiologi- cal and developmental activities; on the other, it is the integration of these activities over time and space that ultimately produces the roots, shoots, and flowers that constitute the mature, reproductive plant. The remaining parts of this section will concentrate on the development of the plant cell. This subject has a strong biophysical component, because for the cell to grow, the cell wall, a strong structure, must yield to high pressure-approximately six times that in a pressure cooker. Most of the controls mentioned above have their ultimate effects through some modification of the biophysics of the plant cell wall. Thus, a portrayal of this subject is a convenient format for the illustration of the unique features of plant development. Plant Growth and Morphogenesis Are Dominated by the Plant Cell Wall Plant cells cannot move appreciably relative to each other; to cover distance they must grow across it. This they do in impressive fashion: more than 100 meters in the height of a redwood tree, scores of kilometers of root length in a typical prairie grass. Much of plant development, and thus many issues in agri

OCR for page 140
172 OPPORTUNITIES IN BIOLOGY culture and forestry, center on how fast and in what direction this growth occurs. A blade of grass, an orchid flower, or a ponderosa pine tree all achieve their configurations by growth and division of walled cells. The two features of major significance in plant cell growth are rate and direction. These two features are controlled, for the most part, by the two main structural components of the cell wall. A strong fibrous component is made of cellulose microfibrils, the orientation of which controls direction. The fibrils are embedded in the second component, a gellilce matrix of other carbohydrate and some protein; the properties of this matrix control rate. The structure is thus like that of the hull of a fiberglass boat. The fibers provide great strength; the matrix distributes stresses equitably so that they do not concentrate at any point to break the fibers. Capitalizing on this principle, plants have come to physically dominate much of the earth's surface. Control of Growth Rate Is Influenced by Hormones It has been known for more than 50 years that a hormone, made in a localized region of a plant, can promote elongation in other parts. The compounds attain and gibberellic acid play major roles in this process, and their structures, synthetic pathways, and modes of degradation are relatively well known. The investigation of the precise ways in which they affect growth, however, is one of the central problems of plant development. Several major features of this process are already clear. The wall extends in response to turgor pressure inside the cell. The cell is able to use osmosis to inflate itself to high pressure with water. The three potential physical controls on the rate of expansion are the turgor pressure itself, the yielding properties of the wall, and the ability of water to enter the cell rapidly. Two of the three possibili- ties have largely been eliminated. There is no support for the idea that plant growth hormones increase turgor directly, and the rate of entry of water is clearly not limiting. In view of these relations, the hormones must be able to cause the wall to stretch or yield. How do the hormones "soften" the wall? The simplest conception of the process views the wall matrix as viscous, like tar or taffy; in such a model, the hormone would simply act by reducing wall viscosity. Such a simple physical model cannot apply in any direct way, however, because the expansion process requires continuous metabolism: For example, inhibitors of oxidative metabm lism stop the growth process as soon as they arrive at the growing site. The con- clusion is that plant growth hormone action is complex, either coupled to the synthesis of new compounds or to the activation of special metabolism. Plant growth, which depends on the growth of cells, is a self-stabilizing and therefore complex process. If the wall-softening processes get out of hand, the cell will burst and die. If synthesis, which must in the long run compensate for the stretching of the wall, is excessive, the wall will be too thick ever to elongate again. One can thus look for a control circuit with both physical and biochemical

OCR for page 140
DEVELOPMENT 173 components. The full clarification of this circuit constitutes one of the major research opportunities in plant growth and development, with important implica- tions for progress in agriculture and forestry. New methodology promises to make this easier. A remarkable "pressure probe" device enables one to measure turgor pressure continuously in the growing cell. This allows the intimate study of metabolic action on the wall's physical properties because changes in wall properties bring about changes in pressure. Understanding the detailed mechanism by which hormones stimulate rate of growth and the way certain wavelengths of light inhibit it are major prerequisites to the understanding of plant growth. Control of Direction of Growth Depends on the Direction of Cellulose Synthesis Control of the shape of plant cells can be achieved by the highly localized control of growth rate, as is the case for some unicellular structures such as root hairs and pollen tubes. The cells in multicellular plant organs, such as roots, stems, and leaves, however, grow throughout their length and hence change shape by a different method. These cells are directionally reinforced, the cellulose microfibrils in their walls lying transverse to the cylindrical axis of the cell. The cell resembles a barrel made of hoops, with the body of the cell extending through the center of the hoops. Without such directional reinforcement, turgor pressure would swell the cells into spheres. Such a pattern can be explained only by cellulose synthesis in a particular direction (as opposed to random synthesis), and this directionality in turn determines the shape of a plant. The shoot attains its height and the root its depth through the directed extension and periodic division of transversely reinforced cells. The control of direction of cellulose synthesis is achieved by means of cytoplasmic microtubules that lie just inside the plasma membrane of the cells. Any depolymerization or disruption of the alignment of these microtubules will randomize cellulose alignment. Understanding this relation has made possible an important distinction between cellulose synthesis as such, which continues during disruptive treatments, and the control of the direction of this synthesis, which is attributable to alignments in the cytoskeleton of the cell around which the synthe- sis is taking place. Progress has also been made in visualizing the synthesis of cellulose. Parallel strands of cellulose polymer seem to coalesce into microfibrils as they emerge from rosettes of protein molecules in the plasma membrane. The involvement of such complex structures at the site of production may explain why such cellulose production has not yet been achieved in vitro. A current major research opportunity is to understand the nature of the coupling between microtu- bule direction and the control of the direction of the spinning out of new microfi- brils. This connection is important in understanding the geometry of plant bodies. Microtubule orientation, which is central to cell wall formation, is also involved in certain plant hormone responses. For example, ethylene, a gas that helps fruit ripen and acts as a plant hormone, causes many organs to swell. Such

OCR for page 140
174 OPPORTUNITIES IN BIOLOGY swelling involves a rotation of cellulose synthesis in cell walls from the transverse to the longitudinal, thus bringing on a corresponding rotation in the direction of growth of the cells. Whereas ethylene causes cells to become less markedly cylindrical, gibberellin, enhances their cylindrical form, apparently by improving cellulose microfibril alignment. When gibberellin is active, the microtubules are more numerous and better aligned. The connection between hormone presence and microtubule influence on the pattern of cell-wall growth is a key factor in understanding plant development and one that has come into prominence only recently. The influence of hormones on microtubule orientation and hence on cell form and their subsequent influence on growth rate are of considerable practical impor- tance for agriculture and forestry. For example, one class of weed killers contains growth hormone analogues that disrupt relative growth rates so that undesirable plants will die; another class of weed killers destroys weed seedlings by interfer- ing with their directional growth. In the future, further mastery of these processes, combined with the techniques of genetic engineering, will have the potential to produce crops of improved form, thus enhancing productivity directly. The Development of PLant Organs Plant Meristems May Be Viewed as Developmental Engines As the study of the extension of the cell wall dominates studies of plant development at the cellular level, so the study of cyclic or continuous morpho- genesis in well-defined meristems dominates the study of whole-plant develop- ment. Meristems are zones of continuous cell division located at the tips of stems and roots in plants. For example, the shoot tip returns repeatedly to the same configuration while continuously producing leaves and additional stem. Such a tip differs from all mechanical analogues since it continually incorporates new material in its products by forming new cells. Three characteristics~ersistent activity, the production of new organ structure, and ever-renewed cell composi- tion-must be combined in any coherent theory of meristem development. The shift to flowering occurs when the vegetative meristem (producing a consistent leaf pattern) of a plant becomes an embryonic inflorescence. Within these embryonic flowers, the internodes are greatly compressed. Unlike the consistency seen with leaf production, the floral organs change character in suc- cessive rounds of meristematic activity. From a central mound of tissue come sepals, petals, stamens, and carper& the members of the four whorls that make up a complete flower. The ways in which the interaction of light and hormones bring about this conversion are explored in Chapter 11; their cellular details remain poorly understood. For example, why do stamens arise at certain places on the meristem? The application of various techniques that have recently become available promises to shed light on these processes in the near future.