Heredity and Development: Second Edition (1972)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

HEREDITY AND DEVELOPMENT: SECOND EDITION 256 12 Differentiation The Problem of Differentiation. The main problem of embryology is this: How, in the course of development, does a cell of one type change into other types of cells? Cell division divides the fertilized ovum into the many cell types of the adult: eye cells, heart cells, kidney cells, nerve cells, and so on— all traceable to an identical beginning. How can the fact of differentiation be made to conform to the fact of mitotic cell division, a process that seems to give daughter cells identical with each other and with the parent cell? Development has fascinated human beings for a very long time. The growth of a huge tree from a tiny seed, the conversion of an egg into a chick, and, of course the development of man himself—all the more mysterious because the early stages are hidden from view. How can an egg, apparently formless and often so small as to be nearly invisible, have the potential for producing a large and complex organism? Interest in this problem is not new. Aristotle was fascinated with development. He studied many sorts of embryos, especially chick embryos, and his descriptions and speculations were not improved upon for 2000 years. PREFORMATION AND EPIGENESIS We can begin our study of embryological concepts at the close of this long and sterile period. Much of the speculation of the seventeenth and eighteenth centuries concerned two rival concepts of development, preformation and epigenesis. Preformation, as the term implies, means that all the parts of the adult, including the most minute ones, are already perfectly formed at the very beginning of development. Epigenesis means 

HEREDITY AND DEVELOPMENT: SECOND EDITION 257 that the adult parts are not present at the beginning, but are developed during embryonic life. Preformation. Preformation was the generally accepted doctrine during the seventeenth and eighteenth centuries. The basis for this belief was largely philosophical and, to a lesser extent, theological. In addition there were some fascinating ‘observations’ that seemed to support the concept of preformation. Some of the greatest scientists of the time, such as Swammerdam, Malpighi, Leeuwenhoek, Leibnitz, Reaumur, Spal- lanzani, and Bonnet, were preformation- ists. They were divided into two schools: the ‘ovists,’ who believed that a tiny pre- formed body was present in the ovum, and the “spermatists,’ who believed that a tiny body was present in the sperm. Swammerdam, an ovist, believed that the animal hemisphere of the frog’s egg con- tained a tiny frog. This subsisted on the food material of the vegetal hemisphere. Embryonic development was simply the increase in size of this tiny frog. Others thought they saw little chickens in the unincubated hen’s egg. The spermatists reported tiny animals in sperm. Figure 12–1 shows what one observer thought he would find if he could see the inner detail of the human sperm. Many others claimed to confirm him, not only on human sperm but on animal sperm as well. There is no problem of differentiation for the preformationists: the adult structures are already differentiated at the beginning of embryonic life; development consists only of growth. The theory of preformation had one inter- esting corollary. Let us adopt the sperma- tist position and consider the tiny creature curled up in the sperm head of Fig. 12–1. Let us further suppose that this gametic 12–1 Homunculus in human sperm Lilliputian is a male. If so, (from Hartsoeker, Essai de dioptrique, Paris, 1694, p. 230). 

HEREDITY AND DEVELOPMENT: SECOND EDITION 258 12–2 (legend on facing page). his testes must be fully formed and contain sperm. These sperm would con- tain fully formed creatures as well. They would be the potential children of the homunculus. These children would contain the grandchildren, and the grandchildren would contain the great-grandchildren, all encased like a set of Russian dolls. Ridiculous as this may seem today, speculations of this sort were made in all seriousness. It was stated by some that Adam or Eve (depending on whether the author was a spermatist or an ovist) must have had the ‘seeds’ for all future generations of mankind in his or her 

HEREDITY AND DEVELOPMENT: SECOND EDITION 259 12–2 Four stages in the development of the chick (from Malpighi, ‘De Formatione pulli in ovo,’ in Opera Omnia, Scott and Wells, London, 1686). gonads. The human race would come to an end when the supply of succes- sively encased homunculi was exhausted. It was probably a philosophical difficulty that made men believe in the theory of preformation when observation might have convinced them of epi- genesis. How was one to understand that something entirely new could appear during development? How could the heart, or the brain, suddenly appear where only formless protoplasm existed before? If the egg of the frog was a structureless body detached from any influence of the parent, how was it able to develop into an exact replica of its species? Why did it not form a toad, or a fish, or an elephant? Clearly something ‘preformed’ must be trans- mitted from parent to offspring, or the frog’s egg would not develop into a frog. No doubt it was easier for the embryologists to imagine that structure, rather than anything else, was transmitted. For this reason it was assumed that the earliest embryo must be a miniature adult. As an example of how reason can lead one to deny what his senses reveal, we should say a few words about Marcello Malpighi (1628–94), an outstand- ing seventeenth-century Italian biologist. (Malpighi made many discoveries. He was, for example, the first to observe capillary con- 

HEREDITY AND DEVELOPMENT: SECOND EDITION 260 nections between arteries and veins. William Harvey, who is generally cred- ited with the theory that blood circulates, assumed that connections must exist, but he did not see them.) Malpighi was a preformationist, yet his work on the hen’s egg showed that development was epigenetic. The figures in 12–2 are his drawings of the developing chick. In the unincubated egg there is a tiny embryo. (In the chicken, fertilization is internal; some development occurs before incubation begins.) In increasingly older embryos new struc- tures make their appearance, and gradually a chicken-like creature is pro- duced. Malpighi believed that this epigenetic phenomenon was an illusion. To him, the structures that made their appearance were there all the time; he was just not able to see them. If Malpighi had not let his beliefs overrule his observations, posterity would have hailed him for presenting the first well- documented case for epigenesis. Malpighi should not be criticized for disbelieving what he saw. Scientists throughout the ages have had to face conflicts between preconceived beliefs and observations. Not infrequently the observations are made to fit the pre- conceived beliefs. Sometimes this has been proved, by subsequent events, to be the correct thing to have done. For example, consider Flemming’s belief, mentioned in Chapter 2, that chromosomes are constant cell structures. He held this view in spite of the fact that the chromosomes seemed to disappear between successive divisions. At other times the acceptance of beliefs con- trary to observation has retarded the progress of science for years. In Malpighi’s case, observation should have triumphed over reason. It did not. We must realize, of course, that a verdict of this sort can be given only in ret- rospect. It is not intellectually dishonest to fit observations to ideas: a scien- tist must seek to fit observations into reasoned order. Epigenesis. In spite of an almost universal belief in preformation, there were always men of renown who thought that development was epigenetic. They believed that adult structures were absent from the early embryo, and that they made their appearance during the course of embryonic life. This was the belief held by Aristotle and, two millennia later, by the English biolo- gist William Harvey (1578–1657). In 1759 the dissertation of Caspar Friedrich Wolff (1733–94), a German zoologist, was published, the most careful description of the developing chick that had been made. Wolff believed in epigenesis, and the observations reported in his dissertation appeared to confirm his belief, but during his life- time he was unable to convince many of his contemporaries. In fact it was not until the early part of the nineteenth century that a majority of biologists finally accepted the concept. The 

HEREDITY AND DEVELOPMENT: SECOND EDITION 261 careful observations of Wolff and those who followed him finally convinced embryologists that new structures do make their appearance in development. Epigenesis is, of course, the view that we hold today. The study of the pho- tographs of frog embryos (Figs. 10–1 to 10–25) will indicate that the fertil- ized ovum is not an adult in miniature. Development begins in a relatively structureless and homogeneous mass, and gradually the organs and parts dif- ferentiate to produce the adult body. The problem of differentiation, which would be non-existent if preformation were true, returns in full force and must be explained. MOSAIC AND REGULATIVE DEVELOPMENT The Mosaic Theory of Development. The next major attempt to solve the problem of differentiation was made by the German embryologist Wilhelm Roux (1850–1924) in the last two decades of the nineteenth century. He was working at a time when spectacular discoveries of chromosome movements in mitosis and meiosis were being made (Chapter 2). Roux attempted to apply the cytological information to embryological problems. According to him, the zygote nucleus contains the determinants for differentiation. These determinants were localized in the chromosomes, and during cleavage they were parcelled out to separate cells. Finally, each cell would be left with only the determinants of a single sort, such as heart-cell determinant, muscle-cell determinant, and so forth. The theoretical biologist, August Weismann (1834–1914), also had a prominent role in the development of this theory. Cleavage and the Polarity of the Frog Embryo. Roux was led to this hypothesis by some observations he made on early development in the frog. Shortly after fertilization, and before first cleavage, some of the black pig- ment of the animal hemisphere, near the equator, disappears, producing the gray crescent. The plane of first cleavage cuts through the center of the gray crescent. Still later the dorsal lip of the blastopore forms in the center of the area where the gray crescent had been. The anus forms where the blastopore closes. Thus the position of the gray crescent corresponds to the posterior end of the embryo. Since the plane of first cleavage runs through the gray cres- cent, it can be looked upon as dividing the embryo along what will be the future midline. Thus one of the two cells formed will become the right side of the body and, the other, the left side. Roux believed that the first mitotic division of the embryo resulted in the determinants for all structures of the right side going into the right cell and those for the left side going into the left cell (Fig. 12–3). 

HEREDITY AND DEVELOPMENT: SECOND EDITION 262 The plane of the second division is at right angles to the first, cutting across the body of the future embryo. The four cells will represent the right anterior, left anterior, right posterior, and left posterior sections of the future embryo. At this second division, Roux believed, the determinants for the four regions of the future embryo are segregated. The simple diagram of Figure 12–3 will show how the determinants were thought to be allocated to each cell by the mitotic division. Continuing cell division results in a continuing segregation of determinants. Finally, each cell will contain a specific determinant, which will be responsible for its differentiation into a specific type of cell. Roux’s Test of the Mosaic Theory. Roux sought to test this hypothesis in one of the first experiments ever performed on an embryo. If the hypothesis is true, one of the cells of the two-cell stage should contain the determinants for the left side of the body, and the other cell the determinants for the right side of the body. Now, if one of the cells is killed, the embryo should have the determinants for only one side of the body. It should develop into a half- embryo. Roux performed this experiment by killing one of the two cells with a hot needle. The uninjured cell cleaved as though it were still part of the entire embryo. 12–3 Roux’s theory of the segregation of determinants during cleavage. LA is the determinant for the left anterior quadrant of the embryo. RA, LP, and RP are determinants for the other quadrants. Gray crescent area stippled. 

HEREDITY AND DEVELOPMENT: SECOND EDITION 263 12–4 Half-embryos obtained after killing one cell of the 2-cell stage. a is a blastula with the dead cell adjacent to the living half-embryo. b is an early neurula. The dead cell has been cast off (from Roux, 1888. Virchow’s Archiv. 114:113). It formed a half-blastula (Fig. 12–4), underwent an abnormal gastrulation, and produced a structure that was interpreted as a half-embryo. A single neu- ral fold was formed, and the mesoderm was present only on the uninjured side. Roux’s experiment seemed to be a dramatic proof of his theory of the seg- regation of determinants during cleavage. The uninjured cell formed only that part of the embryo it would have formed in an uninjured embryo. Roux interpreted this to mean that each cell was capable of forming only the struc- tures it would in normal development. One could imagine the embryo to be a mosaic of cells, each cell capable of producing only a specific part of the adult body. Mosaic Development in Other Embryos. Other investigators, experiment- ing with embryos of many kinds of animals, obtained results that seemed to confirm Roux’s hypothesis. In some annelid worms it was found that cleav- age is so precise that it is possible to trace the lineages of single cells. For example, it was observed that one cell, which forms at the sixth cleavage, gives rise to all of the mesoderm. In some species of tunicates (marine ani- mals related to the vertebrates) the eggs have a pronounced pigment pattern. This enabled the observer to trace the various regions during development, in much the same way as Vogt did many years later in his experimentally stained amphibian embryos. It was found that each cell of the early embryo gives rise to specific structures in later embryo. Embryos of annelid worms, mollusks, tunicates, and several other groups were found to behave in this manner. They were spoken of as having mosaic development. The embryo was regarded as an association of independent cells each developing along a predestined path to form a specific part of the adult. 

HEREDITY AND DEVELOPMENT: SECOND EDITION 264 The Reinterpretation of Roux’s Results. The novelty of Roux’s theory and its experimental verification convinced many embryologists that an under- standing of differentiation was at hand. It was not long, however, before complications and exceptions were noticed. It was found that the results Roux obtained after killing one cell of the two- cell stage embryo were due largely to the presence of the dead cell. Whether for mechanical or other reasons, the dead cell seemed to prevent the living cell from rounding up and producing a whole embryo. If one of the cells was removed by sucking it out with a pipette, the remaining one produced an entire embryo, which differed from the normal only in being smaller. In another experiment a fine thread was tied around the embryo at the two-cell stage and tightened until the two cells were separated. Each of the cells pro- duced a normal embryo. One could conclude from these experiments that there was no evidence of a segregation of determinants as postulated in Roux’s theory. Regulative Development. The early embryo of the frog is not the mosaic of independent parts as Roux had believed. Instead, its development is more like that of the sea urchin. In the sea urchin it was found that the cells of the two-cell stage could be separated and each cell would give rise to a normal embryo. The same was true at the four-cell stage. In contrast to the mosaic embryos, in which each cell could form only the part it would in normal development, the sea-urchin embryos were said to be regulative. Isolated cells of the regulative eggs could adjust to the new situation and produce a whole embryo. The conclusion was reached that there are two main patterns of develop- ment, mosaic and regulative. This was interesting but it shed no light on the fundamental causes of embryonic differentiation. Roux’s hypothesis was shown to be inadequate and many years were to pass before there was a use- ful substitute. THE ORGANIZER THEORY The next major advance in the study of differentiation was made by the Ger- man embryologist Hans Spemann (1869–1941) and others in experiments on the differentiation of the amphibian nervous system. Their studies led to the organizer theory, which is the most important embryological concept pro- posed during the first half of the twentieth century. Formation of the Neural Tube in the Amphibian Embryo. If we exam- ine the cells of an amphibian embryo in the late blastula stage we find that they are essentially the same in appearance throughout the entire embryo. There is a gradient of increasing cell size extending from the 

HEREDITY AND DEVELOPMENT: SECOND EDITION 265 animal to the vegetal pole, and the concentration of yolk granules in the indi- vidual cells is subject to variation, but beyond this there is little to suggest the widely divergent destinies that will befall the cells of different regions. The conversion of the single-celled zygote into a many-celled blastula is brought about by cleavage, with little or no visible differentiation of the cells: the cells just get smaller. During gastrulation the cells become rearranged and the three germ layers can be distinguished, but even at this time there is little difference among the ectoderm, mesoderm, and endoderm cells. Subse- quently the slow process of cellular differentiation results in various visibly different cell types, such as muscle, gland, and nerve, that make up the tissues and organs of the embryo. We have previously learned that the nervous system is one of the first organ systems to make its appearance. Observations of the living embryo give us considerable information about its formation. At the end of gastrula- tion the embryo becomes flattened on the dorsal side. This flattened area is the neural plate. Next the neural folds appear as ridges along the periphery of the neural plate (Fig. 10–17). These folds move toward the mid-line and fuse along their crests (Figs. 10–18, 10–19). In this manner the neural plate is con- verted into a tube lying beneath the now continuous ectoderm in the dorsal part of the embryo (Fig. 10–21). Observations of the living embryo could be supplemented by the study of sections prepared by the usual histological techniques. From these we could obtain information on the changes occurring within the embryo. We would find that by the time the neural plate is formed, gastrulation movements have brought a sheet of mesodermal cells into position beneath the neural plate (Figs. 11–6 to 11–8). Somewhat later these mesodermal cells will form the notochord and myotomes. Repeated observations would show that the neural plate and tube are always formed in the same part of the embryo. In normal development, the ectoderm cells situated on the side of the embryo above the blastopore and their descendants produce these structures (Fig. 11–4). The appearance of the neural plate in a constant position suggests that the cells that form it differ in some way from other ectodermal cells. They alone develop into the neural plate, while the remaining ectoderm cells produce the epidermal covering of the body. It is possible to trace the positions of the presumptive neural plate and the presumptive epidermis cells back to the early gastrula, as was done by Vogt (Chapter 11). In all probability, we could even trace the presumptive regions back to the early cleavage stages. Hypotheses of Neural Tube Formation. What is different about the portion of the ectoderm that will form the neural tube? Why does it 

HEREDITY AND DEVELOPMENT: SECOND EDITION 266 and no other part of the ectoderm form this structure? Observations of a nor- mally developing embryo cannot answer such questions. We can only attack the problem by experimentation. Yet what experiment can we perform? First we must formulate a question that is precise enough for us to seek a definite answer. We already know that the group of cells that occupies the presump- tive neural tube region of the early gastrula will, in later stages, form the neu- ral plate and, still later, the neural tube. Two alternative hypotheses to explain this phenomenon could be suggested: Hypothesis 1. The presumptive neural tube cells of an early gastrula possess an inherent capacity to form neural tissue. That is, they have within themselves all that is necessary to differentiate into a neural tube. Hypothesis 2. The presumptive neural tube cells of an early gastrula do not pos- sess an inherent capacity to form neural tissue. Influences from outside the per- sumptive neural tube area are necessary for differentiation of a neural tube. Tests of the Hypotheses. These hypotheses are formulated in such a manner that they can be tested. Let us begin with the first hypothesis. If the presump- tive neural tube cells possess within themselves all that is necessary for neu- ral tube differentiation, we can make the following deduction: The presumptive neural tube cells should be able to differentiate into a neural tube if they are separated from the remainder of the embryo. This separation can be accomplished in several ways. Experiment 1. Separation of the presumptive ectoderm from the remainder of the embryo in exogastrulation. This experiment was performed by Johannes Holtfreter (born 1901). He removed the membranes from an early gastrula, oriented it with the animal hemisphere down, and let it develop in a solution with a salt concentration higher than in pond water. Under these con- ditions gastrulation movements were quite abnormal. The presumptive ecto- derm cells did not move down over the vegetal hemisphere, but tended to pull away from the remainder of the embryo. The result was a dumbbell-shaped embryo known as an exogastrula. In extreme cases the presumptive ecto- derm cells formed a ball connected by only a thin strand of cells with the pre- sumptive endoderm and mesoderm (Fig. 12–5). Further development of the two parts was very different. The cells of the presumptive endoderm and mesoderm were able to differentiate into a heart, muscles, parts of the alimentary canal and other organs 

HEREDITY AND DEVELOPMENT: SECOND EDITION 267 12–5 Holtfreter’s exogastrulation experiment. a is an exogastrula and b is a diagram of the parts (from Holtfreter, 1933. Biologischen Zentralblatt, 53:404). normally formed from these two layers. In marked contrast, the presumptive ectoderm cells remained undifferentiated. No trace of a nervous system was found. This indicates that the presumptive neural tube cells do not possess an inherent capacity to form a neural tube. Hypothesis 1 is false. We must not be too quick to accept this conclusion, however. Perhaps the manipulation of the embryo injured the presumptive neural tube cells in such a way as to prevent them from differentiating into neural tissue. If so, the experiment is not a valid refutation of the hypothesis. This explanation is improbable, however, since the presumptive endoderm and mesoderm showed a considerable amount of differentiation after being subjected to the same experimental conditions. 

HEREDITY AND DEVELOPMENT: SECOND EDITION 268 Experiment 2. Explanation of presumptive neural tube cells. This experi- ment was also performed by Holtfreter, and several other investigators. Pieces of the blastocoel roof of an early gastrula can be cut off and cultured in dilute salt solutions (Fig. 12–6). They remain alive for many days. This exper- imental technique is known as explantation and the pieces as explants. Explants from the presumptive neural tube and presumptive epidermis areas of an early gastrula fail to differentiate into neural tissue. They produce noth- ing more than a simple epidermal type of cell. Results of Experiment 2 likewise suggest that Hypothesis 1 is false. We can only hope that neither the cutting nor the culture conditions injured the explant in some manner that would prevent the cells from revealing their full potential. Such a possibility can partially be ruled out, since explants from some other parts of the embryo are able to differentiate. If this experiment is repeated at the end of gastrulation the results are dif- ferent (Fig. 12–6). Explants of presumptive epidermis form only epidermis but explants of presumptive neural tube cells differentiate 12–6 Explantation of presumptive epidermis and presumptive neural tissue in early (above) and late (below) gastrulae. (Refer to figs. 11–5 and 11–8 for full labels.) 

HEREDITY AND DEVELOPMENT: SECOND EDITION 269 into neural tissue. Some important change has occurred in the presumptive neural tube cells during the interval between the beginning and end of gastrulation. The two experiments have given similar answers. If the presumptive neu- ral tube cells are removed from an early gastrula, either by exogastrulation or explanation, no neural tissue is formed. If these results can be accepted, and it seems probable that they can, the presumptive neural tube cells in an early gastrula do not possess an inherent ability to form neural tissue. Thus, Hypothesis 1 is false. Experiment 2 did show, however, that the presumptive neural tube cells of the late gastrula possess an ability to differentiate into neural tissue if explanted. Some change must occur in the interval between the early gastrula and late gastrula stages. This change does not occur during the development of the ectoderm in exogastrulae or in explants. Thus, it is likely that an influ- ence from some non-ectodermal part of the embryo is responsible for this change in the presumptive neural tube cells. This is our second hypothesis, which was stated thus: ‘The presumptive neural plate cells of an early gas- trula do not possess an inherent capacity to form neural tissue. Influences from outside the presumptive neural plate area are necessary for differentia- tion.’ One deduction that we might make from this second hypothesis is: The neural tube should form in the same position, relative to the non- ectodermal parts of the embryo, no matter how the presumptive ectoderm is oriented. This can be tested as follows: Experiment 3. Rotation of the animal hemisphere of an early gastrula. This experiment was performed by Spemann. It will be recalled that the pre- sumptive neural tube cells occupy that portion of the animal hemisphere near- est to the dorsal lip (Fig. 11–4). The presumptive epidermis cells are on the opposite side of the embryo. If the upper portion of an early gastrula is cut off, rotated 180°, and put back, the cells will come into new relations with the remainder of the embryo (Fig. 12–7). The presumptive epidermis will now be closer to the dorsal lip and the presumptive neural tube will be on the far side of the embryo. The development of the embryo on which the operation was per formed continues as though nothing has happened. The neural folds are formed in the normal relation to the blastopore. This means that the neural folds of this experimental embryo are formed largely from presumptive epidermis, and that the epidermis is derived almost entirely from the presumptive neural tube cells. The results of Experiment 3 indicate that Hypothesis 2 may be cor- 

HEREDITY AND DEVELOPMENT: SECOND EDITION 270 12–7 Rotation of the animal hemisphere. (Refer to fig. 11–4 for full labels.) rect. The differentiation of the ectoderm appears to be greatly influenced by the ventral part of the embryo. Furthermore, the constant relation with the blastopore suggested to Spemann that the cells invaginated at the dorsal lip might be the stimulus for neural differentiation. In normal development these cells form the roof of the archenteron, which is immediately beneath the cells that will form the neural tube. If we regard this position as significant, we might restate the hypothesis more precisely: ‘The presumptive neural plate cells of an early gastrula do not possess an inherent capacity to form neural tissue. Differentiation into neural tissue is the result of stimulation by the roof of the archenteron.’ If this hypothesis is true, we might make the following deduction: If the doral lip cells are removed from one embryo and grafted onto another, and if they are able to invaginate under the ectodermal cells in this new posi- tion, we would expect these ectodermal cells to form neural tissue. This deduction can be tested by an extremely delicate micro-surgical experiment. Experiment 4. Transplantation of the dorsal lip. This experiment, reported by Spemann and Hilda Mangold in 1924, is one of the classics 

HEREDITY AND DEVELOPMENT: SECOND EDITION 271 of embryology. The embryos of two species of salamanders were used. In one species the embryos were nearly white and in the other they were brown- ish. A small piece of tissue was removed from the dorsal lip region of one embryo (Fig. 12–8). This piece of tissue was then transplanted to an early gastrula of the other species in a position 180° from the host’s dorsal lip. (The embryo from which the dorsal lip was removed is called the donor. The embryo to which the transplant is made is called the host.) Since the host and donor tissues were of different colors they could be distinguished. Spemann and Mangold’s operation did not seem to affect the process of gastrulation in the host. Invagination occurred at the host’s blastopore. Of greater significance was the fact that invagination occurred where the trans- planted dorsal lip was placed. The donor cells invaginated through this sec- ondary blastopore and produced a tiny archenteron. At the time the host’s neural folds were forming, neural folds also appeared above the region where the donor tissue had invaginated. These neural folds were composed of host cells. The transplanted dorsal lip had exerted a profound influence. Host cells, which in the course of normal development would have produced epidermis, were changed in such a way that they formed neural folds. Later these neural folds closed. When the embryos were examined in sections it was foun d that the transplanted dorsal lip had stimulated the formation not only of a nerve tube but other structures as well. In some cases almost an entire normal embryo arose in the region where the transplant was made. Interpretation of the Results. Spemann and Mangold, recognizing that the dorsal lip was of great importance in the formation of embryonic structure, called it the organizer. This organizer was postulated to act on other cells to alter their course of development. This action is spoken of as induction. This experiment, and many others, has shown that the neural tube of a normal embryo is formed under the influence of the organizer. At the begin- ning of gastrulation, the organizer region consists of cells above the dorsal lip, in the area corresponding roughly to the presumptive notochord of Vogt’s map (Fig. 11–4). This region invaginates to form the roof of the archenteron. The roof of the archenteron then induces the overlying ectoderm cells to form a neural tube. We now have the theoretical basis to interpret the results of the experi- ments on explanting pieces of presumptive neural tube cells. When these cells are explanted at the early gastrula stage they form epidermis but no neu- ral tubes. When the presumptive neural tube cells are explanted at the end of gastrulation they are capable of forming neural 

HEREDITY AND DEVELOPMENT: SECOND EDITION 272 12–8 Dorsal Lip Transplantation experiment of Spemann and Mangold. a is a diagram of the operation. b is a lateral and dorsal view of a neurula with the secondary embryo. c is an older stage. d is a cross section showing the structures of the primary and the secondary embryos. (b, c, and d modified from Spemann and Mangold, Archiv für Mikroskopische Anatomie und Entwick- lungsmechanik, 100:599. 1924.) 

HEREDITY AND DEVELOPMENT: SECOND EDITION 273 tubes. The explanted early-gastrula presumptive neural tube cells did not form neural tubes because the organizer had not acted upon them. At the end of gastrulation, on the other hand, the organizer area would be in the archen- teron roof and the overlying presumptive neural cells would have been induced. The early-gastrula presumptive neural tube cells are said to be undeter- mined with respect to their ability to form a neural tube. Once the organizer has acted upon them they are determined. Said in another way, determination results from induction. After determination has occurred differentiation is possible. Both induction and determination are invisible biochemical events. They are tested by explantation and other experimental techniques. If a piece of tissue is explanted in a suitable medium and it does not differentiate in a specific way, it is said to be undetermined. If it differentiates after explanta- tion, this shows it was determined at the time it was removed from the embryo. These various experiments have enabled us to choose between the two hypotheses concerning the formation of a neural tube, namely, whether (1) the presumptive neural tube cells of an early gastrula have an inherent ability to form a neural tube, or whether (2) these cells must be acted on by other parts of the embryo in order for them to differentiate in their specific manner. All the evidence given indicates that the second hypothesis is correct. At the beginning of gastrulation, the presumptive neural tube cells do not possess an inherent ability to form a neural tube. The only inherent ability they possess is to form simple epidermal cells. The ability to form a neural tube results from the induction of the presumptive neural tube cells by the organizer in the archenteron roof. At the beginning of gastrulation the entire presumptive ectoderm, whether it is in the presumptive neural tube or presumptive epi- dermis region, is identical in developmental capabilities. One portion forms a neural tube because the organizer acts on it; the remainder forms epidermis because the organizer does not come in contact with it. It was soon found that there is not just one organizer but many. Organizers are known for the mouth, heart, lens, otic vesicle, pronephros, and for other structures as well. We shall consider one example of these secondary orga- nizers, so called because they are produced in structures which have been formed under the influence of the primary organizer of the archenteron roof. Formation of the Lens. Shortly after closure of the neural folds, paired outgrowth, the optic cups, appear in the ventro-lateral portion of the brain. They grow outward until they reach the epidermis. The epidermal 

HEREDITY AND DEVELOPMENT: SECOND EDITION 274 cells adjacent to the optic cups then differentiate into the lenses (Fig. 11–13). Vogt showed that the cells that will form the optic cup and the cells that will form the lens occupy very different regions of the early gastrula (Fig. 11–4). The optic cup area is in the presumptive neural plate, while the lens area is in the presumptive epidermis. The two areas are brought into close association by the complex movements of gastrulation and neural tube formation. Exper- iments have shown that, in some species at least, the optic cup contains the inductor which is the stimulus for the formation of the lens. Let us experiment with an embryo in which the optic cups are beginning to form by making a slit in the epidermis over the brain. The optic cup on one side can be cut off and the epidermis put back in place. The cut tissues heal in a matter of minutes, and the embryo appears to suffer no injury as a result of the operation. We then allow the embryo to develop. The optic cup on the unoperated side grows out toward the epidermis and the epidermis produces a lens. On the operated side there is no regeneration of an optic cup and no lens is formed by the presumptive lens area of the epidermis. Thus, in the absence of an optic cup, lens differentiation does not occur. This result sug- gests that the optic cup may be acting as an organizer for the lens. Another type of experiment substantiates this hypothesis: Let us remove the optic cup as it is forming, make a slit in the epidermis of the flank region of this same embryo, and push the optic cup into this slit. This slit will heal and the optic cup will then be adjacent to epidermis that would normally form only the outer layer of skin. Under the influence of the optic cup, however, the fate of this epidermis is changed—it differentiates into a lens. The embryo then has a structurally normal eye in the flank. (It is non-functional, however, since the proper nerve connections with the brain are not made.) Obviously, these results are similar to the archenteron roof-ectoderm rela- tionships. In the present case the optic cup is an organizer. It induces lens formation in any ectoderm of the right age with which it comes into contact. In normal development, this is the ectoderm of the head but, as experiments have demonstrated, other ectoderm will suffice. The Nature of the Organizer. Following the discovery of the neural tube organizer, many embryologists sought to learn as much as they could about it. The location of early gastrula cells that were capable of induction was found to be restricted roughly to the presumptive notochord and adjacent mesoderm. Transplants of living cells from other parts of the early gastrula did not induce neural structures. 

HEREDITY AND DEVELOPMENT: SECOND EDITION 275 The most encouraging discovery was the ability of dead organizer tissue to induce. Pieces of the dorsal lip, or of the archenteron roof, could be killed by heat or chemical means and still stimulate the formation of neural tissue in undertermined ectoderm. This suggested that the organizer was a chemical substance, and its stability after heat or chemical treatment was a sign that it might be extractable in a pure form. This early optimism was soon shattered by discoveries that were difficult to interpret. The organizer was found to be much more widely distributed than the early experiments on living tissue had indicated. Parts of the gastrula which would not induce neural tissue when alive were found to induce after being killed. Even more perplexing was the finding that adult tissues, such as liver and kidney, could induce embryonic structures. In addition, some organic compounds appear to induce. Some investigators have attempted to purify the organizer, but the results so far obtained have not been very convincing. Some experiments suggest that the primary organizer is nucleic acid. Further advances in our knowledge of the nature of the organizer are await- ing new ideas and new techniques. The Reacting Tissue. In our discussion of induction we have emphasized the organizer. This may have given the impression that the reacting tissue is passively molded by the organizer. This is not the case. The ability of tissue to respond to organizers is limited in a number of ways. Some of these will be mentioned. We have seen earlier that any portion of the presumptive ectoderm can respond to the archenteron-roof organizer by producing a neural tube. The period during which this response is possible is very short. At or about the stage when the neural folds close, the presumptive epidermis will no longer respond to the archenteron-roof organizer. The ectoderm also has a specific period during which it can respond to the optic-cup organizer and produce a lens. The importance of the reacting tissue can be shown in experiments involv- ing tissue of two different species. The mouth regions of a frog larva and a salamander larva differ considerably. The frog larval mouth is bordered by prominent black, horny ‘jaws’ and rows of tiny ‘teeth.’ These jaws and teeth are ectodermal structures that have no relation to the jaws and teeth of the adult. The salamander larva lacks both the ectodermal jaws and teeth. The anterior portion of the archenteron of both frog and salamander induces the mouth region. An interesting experiment can be performed by interchanging frog and salamander ectoderm in the region where the mouth will form. The salamander 

HEREDITY AND DEVELOPMENT: SECOND EDITION 276 12–9 Induction of mucus glands in frog ectoderm transplanted to the ventral side of the head of a salamander embryo. Mucus glands were induced in the frog tissue and mucus is being secreted by them. embryo will then have its mouth region covered with frog ectoderm. The frog embryo will have its mouth region covered with salamander ectoderm. Which type of mouth will form in the two cases? The results of such experiments are clear-cut. The frog tissue on the sala- mander embryo is induced by the salamander mouth organizer to form a mouth. The mouth it forms, however, is of the frog type. In the same way, the salamander ectoderm on the frog embryo produces a salamander mouth. The tissue always responds in accordance with its specific genetic constitution. A somewhat similar situation is encountered with two other structures in the head region of frog and salamander embryos. The frog has a pair of mucus glands on the ventral side near the mouth (Figs. 10–23, 10–25, and 11–14). These are ectodermal structures induced by the underlying tissue. Salamander embryos lack these mucus glands. Instead, they have a pair of balancers (Fig. 12–9), which are also induced by the underlying tissues. If presumptive epidermis of a frog embryo is transplanted to the region behind the mouth in a salamander embryo, the transplant forms mucus glands. The embryo shown in Figure 12–9 is the result of an operation of this sort. The stringy material, which appears below the host’s balancer, is mucus being secreted by the induced mucus glands. This case is of interest, since it shows that the salamander embryo can induce structures which are not a part of its own morphology. One is left with the impression that organizers are general stimuli, and that the end result of their action is modulated by the genetic limitations of the reacting tissue. RECONSIDERATIONS AND CONCLUSIONS The organizer concept has shed some light on the old problem of mosaic and regulative eggs. In many cases, and perhaps in most, the change 

HEREDITY AND DEVELOPMENT: SECOND EDITION 277 from an undertermined state to a determined state is the result of an influence external to the cells being determined. In other words, something of the nature of organizer action is involved. The main difference between mosaic and regulative eggs is thought to be the time at which these external influ- ences cause determination. In the mosaic eggs this occurs when the egg is being formed in the ovary. The developing embryo is, therefore, mosaic from the start. In regulative eggs, on the other hand, determination occurs during development. Modern work in genetics and embryology has put the old controversy of preformation and epigenesis in a new light. Somewhat earlier, we saw that something ‘preformed’ must be transmitted from adult to embryo, otherwise the frog embryo might form a toad, a fish, or an elephant. Observations on developing embryos ruled out the transmission of preformed adult structures. Genetics and embryology have demonstrated that the preformed entities which are transmitted are the genes and an organized cytoplasm. A frog embryo becomes a frog because it receives the genes and cytoplasm that con- trol the development of a frog body. The hereditary basis of development is preformed in the structure of the gametes; the appearance of adult parts is epigenetic. This brief survey has revealed some of the factors responsible for differen- tiation but embryology has not reached the point where we can say that we know in detail why an embryonic cell differentiates in a specific way. We know some of the answers and when we know more we should be able to answer a question of vital interest: ‘Why do normal cells sometimes change into cancerous cells?’ This also is differentiation. Further progress in embryology will depend upon more information con- cerning the manner in which genes act. In the next chapter we shall explore some of the ways in which the recent data of genetics can be applied to embryological problems. Suggested Readings Chapter 9 in the Readings contains an article on mosaic development by E.B. Wilson and one on the organizer by Hans Spemann. There are also additional references. BALINSKY, B.I. 1970. An Introduction to Embryology. Third Edition. Philadelphia: W.B. Saunders. HUXLEY, JULIAN S., and G.R.DE BEER. 1934. The Elements of Experimental Embryol- ogy. Cambridge University Press. Republished by Hafner, New York. SAUNDERS, JOHN W. JR. 1968. Animal Morphogenesis. New York: Macmillan. SAXEN, LAURI, and SULO TOIVONEN. 1962. Primary Embryonic Induction. Engle- wood Cliffs, N.J.: Prentice-Hall. SPEMANN, HANS. 1938. Embryonic Development and Induction. New Haven: Yale Uni- versity Press. Republished by Hafner, New York. 

HEREDITY AND DEVELOPMENT: SECOND EDITION 278 TRINKAUS, J.P. 1969. Cells into Organs. The Forces that Shape the Embryo. Englewood Cliffs, N.J.: Prentice-Hall. WADDINGTON, C.H. 1956. Principles of Embryology. London: George Allen and Unwin. WILLIER, BENJAMIN H., PAUL A.WEISS, and VIKTOR HAMBURGER. 1955. Analy- sis of Development. Philadelphia: W.B.Saunders. WILLER, BENJAMIN, and JANE M.OPPENHEIMER. 1964. Foundations of Experimen- tal Embryology. Englewood Cliffs, N.J.: Prentice-Hall. Classic papers. WILSON, E.B. 1928. The Cell in Development and Heredity. New York: Macmillan. Chap- ters 13 and 14.