Biological Journal Offhe Linnean Society (1990), 39: 109-124. With 3 figures
The evolution of development
LEWIS WOLPERT, F.R.S.
Department of Anatomy and Developmental Biology, University College and Middlesex
School of Medicine, London W lP 6DB
The evolution of development required few new features not already present in the eukaryotic cell,
as exemplified by the cell cycle. Moreover, the protozoa possess many features of sphial
organization and regulation present in metazoan embryos.
The earliest multicellular organism could have been reproduced by a stem cell mechanism or by
fission, the latter requiring cell-to-cell interactions that may have favoured cell-interactions and
regulation. Regeneration can be considered as a meta-phenomenon related to asexual reproduction
and retention of embryonic characters. The origin of embryonic structures like the gastrula may be
accounted for in terms of Haeckel’s ‘Gastrea’ theory. Mechanisms based on selection at the level of
cell lineage are rejected.
It is not clear what selective forces act on development itself, as distinct from the requirement for
reliably producing a functional organism. There is, for example, a major problem why gastrulation
should be so variable in related animals. Selection for rate of development in relation to energy
utilization may play a role. If many variants are neutral this may facilitate the evolution of novelty.
In general terms there is a requirement for a continuity principle for the evolution of each form in
development. Most groups pass through a phylotypic stage with considerable diversity before and
after.
KEY WORDS:-Cell
heterochrony.
cycles
-
evolution
~
ontogeny
-
multicellularity
-
gastrulation
CONTENTS
Introduction . . . . . . . . . . . . .
Beyond the cell
. . . . . . . . . . . .
The evolution of early differentiation and patterning . . .
The evolution of differentiation.
. . . . . . . .
The origin of the gastrula . . . . . . . . . .
. . . . . .
Selection at the level of the cell lineage.
Selection on developmental processes . . . . . . .
The continuity of development: heterochrony and metamorphosis
Induction . . . . . . . . . . . . . .
Regeneration . . . . . . . . . . . . .
Conclusions . . . . . . . . . . . . .
References
. . . . . . . . . . . . .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
109
110
112
113
114
116
118
120
121
122
122
123
INTRODUCTION
The cell is evolution’s most brilliant invention and development is its
triumphant elaboration. Developmental programmes are long and complicated
and pass through morphological stages like that of the blastula and gastrula
109
0024-4066/90/020109+ 16 SOS.OOj0
0 1990 The Linnean
Society of London
110
LEWIS WOLPERT
which give no hint of the nature of the animal which will emerge. How did such
structures evolve? Why is gastrulation or neurulation different in some animals?
What did the cell need to invent in order to develop? These sort of questions are
not usually considered in the recent revival of interest in the relation between
evolution and development (Raff & Kaufman, 1983; Goodwin, Holder & Wylie,
1983). Attention has been mainly directed at how the developmental
programmes could be modified by, for example, heterochrony, and what
constraints development puts on evolution (Smith et a/., 1985). It is not such
issues that are considered here but rather the evolution of development itself.
The evolution of development requires consideration of a set of problems
rather different from, but overlapping with, those usually treated in relation to
evolution and development. The first main problem is, given the eukaryotic cell,
what new processes had to be invented to generate multicellular animals; and
have new processes emerged during evolution, or have they been highly
conserved? The second main problem is to explain the origin of what are
essentially embryonic structures and processes such as radial and spiral cleavage,
blastula and gastrula, and why these vary. This is closely associated with the
third problem, which is what selective pressures act on development itself, but
particularly those that do not relate simply to the reliable generation of a
functioning organism. Finally there are problems associated with the origin of
novel features like, for example, the evolution of the neural crest, metamorphosis,
and the five arms of echinoids: problems which have a stronger overlap with
conventional considerations of evolution and development.
The problems are very difficult ones; not the least difficulty being our
ignorance of developmental processes.
BEYOND THE CELL
If we consider the three basic processes in development-differentiation,
spatial patterning and change in form (Wolpert, 1969)-these are already, it will
be argued, well developed in the single cell. We will assume that the original
eukaryotic cells had all the basic structures that characterize all typical
cells-membranes, nucleus, mitochondria and so on, and in fact it is remarkable
how similar the basic components of all cells are.
The cell cycle can be taken as a paradigm for several of the key processes in
development. Every cell cycle involves a temporal programme of gene activity, a
spatial differentiation that assigns the results of growth to the daughter cells, and
cell motility which ensures that the chromosomes are distributed to daughter
cells and in animal cells brings about cell cleavage by active constriction. The
evidence for a temporal programme of gene activity is well documented (Alberts
et al., 1989) and it does not seem unreasonable to consider the G,, S and G,
stages of the cell cycle, as being homologous with different differentiated cell
states. In support of this view, it has been shown that in the cellular slime mould
the fate of cells is dependent on the stage of the cell cycle from which they enter
the developmental programme (Weijer, Duschl & David, 1984). Cells early in
the cycle exhibit strong prestalk sorting, while those taken later in the cycle
exhibit strong prespore sorting.
Two further features of the cell cycle merit attention. The first involves the
decision whether or not to enter the cycle following mitosis. This decision is
THE EVOLUTION OF DEVELOPMENT
111
closely linked to growth and the nuclear/cytoplasmic ratio. I t also involves some
sort of threshold event. I n more general terms it involves intracellular signalling
and a switch of a kind that is very similar to many developmental processes.
The other feature meriting attention is mitosis and cleavage. Here there is
highly organized spatial patterning within the cell which is also linked to
motility. The development of the plane of cleavage at right angles to the spindle
foreshadows the orthogonality that is fundamental to embryonic development.
In addition, mitosis provides the opportunity for unequal distribution of
components to daughter cells that could be the basis for the origin of divergence
in cell fate and differentiation. This autonomous generation of differences is clearly
present in the division of yeast cells in relation to mating type (Wolpert, 1989a).
It requires little imagination to derive other properties from the primitive cell
for development. Cell adhesion and cell-to-cell signalling may be novel attributes
of multicellular organisms but require little modification of a system in which
there is already a flow of material to the membrane with an external coat. Even
the provision of an extracellular matrix would seem to present little difficulty,
given an endoplasmic reticulum and Golgi. There is, perhaps, one cellular
process that may require a novel evolution in metazoans and that is cellular
memory. Liver cells breed true. The inheritance of the differentiated state
through a cell cycle might require new methods for controlling gene action, as
the inactivation of the X chromosome suggests.
The single cell protozoa show complex patterning with very precise spatial
location of organelles. Many of the mechanisms required for the evolution of
metazoan development are present in the protozoa (Goodwin, 1989), their
presence suggesting that they represent fundamental biological processes that
arise, as it were, easily and naturally in evolution. Protozoa clearly show polarity
in form. The rows of kineties-lines of cilia-suggest that a mechanism for
generating spacing patterns-stripes-is
well developed. In the ciliated protozoa
fission involves considerable reorganization of existing structures. For example,
in Tetrahymena a new oral apparatus, which is normally located a t one end of the
animal, begins to develop near the zone of fission. I n fact, two new cell surface
organizations are formed within what was originally one, the two now being in
tandem alignment along the antero-posterior axis. This reorganization is
analogous to budding in hydra.
The various cellular structures have a well-defined positional relation with
each other. Many protozoa show phenomena of regeneration that appear to be
very similar to those seen in metazoa (Frankel, 1984). This capacity probably
reflects the processes necessary for fission in a manner similar to that to be
discussed for hydroids and planaria. A variety of experiments illustrate that this
regulation of pattern has properties which are capable of being described in
terms of gradients and positional values in a manner similar to that in the
metazoa. Many of the phenomena show striking similarity as in the generation of
mirror-image patterns, discontinuities and intercalation (Frankel, 1984). I t is
tempting to think that whatever mechanism was used by the protozoa for
patterning was also used by the early multicellular organisms.
Thus it is clear that single cell organisms have many of the mechanisms
required for multicellular development. But these mechanisms must now be used
for multicellular patterning. With a mechanism for spacing patterns and
positional fields, many patterns are possible (Wolpert, 1989b).
LEWIS WOLPERT
112
0
I
A
B
Figure 1. Possible modes of development and reproduction in primitive metazoa, made up of just
two cell types, one of which can be thought of being specialized for feeding. In A, development and
reproduction is based on lineage and a stem cell. The dot represents some determinant that
maintains the non-feeding cell as a stem cell. In B, development and reproduction is based on cellto-cell interactions. One of the cells divides and the organism then reproduces by fission, a new
feeding cell developing in just one of the two cells that have split off. This requires both cell
signalling and polarity.
T H E EVOLUTION O F EARLY DIFFERENTIATION AND PATTERNING
The simplest conceivable multicellular organism would consist of two cells
that differ from one another. This hypothetical organism would have manifested
the two key processes of multicellular development, diversity in cell state, or cell
differentiation, and spatial heterogeneity, for it was essential that the two cells be
different. From the beginning there needed to be a mechanism to ensure this
difference and either cytoplasmic differences or a signal could be the basis
(Fig. 1 ) . Both became firmly established. Almost all eggs, those of mammals
providing a major exception, have a well defined polarity due to cytoplasmic
differences, but in most embryos later differences arise not because of
cytoplasmic localization but because of cell-to-cell interactions.
It is a major problem in the evolution of development to understand why some
animals have very well defined cell lineages, the asymmetric behaviour of
daughter cells being a key feature of their development, whereas others rely
largely on cell-to-cell interactions. What does this diversity reflect? Again, for
example what underlies the differences between spiral and radial cleavage?
One possible scenario for thinking about the difference between mosaic and
regulative development is to relate it to asexual reproduction. Asexual
reproduction by budding as in hydra, or fission as in flatworms, would seem to
demand, and does indeed involve cellular interactions. In hydra, for example,
the bud is organized by the tip of the future head (Bode & Bode, 1984). It is
THE EVOLUTION OF DEVELOPMENT
113
attractive to think that such interactions are linked to regulative development
which involves similar interactions. In this connection it is interesting that some
hydroids have a highly regulative development.
One may thus consider two pathways in the evolution of a primitive metazoan
which has but two cell types (Fig. 1) One pathway is to the development of a
two-celled animal which reproduces via a single cell which, like a stem cell, can
divide either symmetrically or asymmetrically. This is the lineage pathway. In
the other pathway, reproduction is by fission, and requires cell-to-cell
interactions to specify the head or end cell.
Absurdly idealized as these processes are, several points are worth noting.
Firstly, there is a placozoan, Trichoplax (Kaestner, 1980), which is essentially
made up ofjust two cell types, reproduces by fission, and also has a sexual cycle.
Secondly, in algae such as Anabaena there are only two cell types and the
differentiation of one of these in the filament is controlled both by cell
interactions and cell division (Wilcox, Mitchison & Smith, 1973). Thirdly,
mechanisms based on cell lineage link changes in cell state to cell division,
whereas those based cell intractions, permit, in principle, the specification of
multiple cell states at one time and also decouple the specification from cell
division. Moreover cell-to-cell interactions provide a far more flexible
mechanism and most important, permit co-ordination between non-adjacent
cells.
One can speculate on the use of gradients in chemical concentration that may
have provided positional information in both protozoa and early embryos. Such
gradients could have been generated by reaction-diffusion type mechanisms.
This, together with a mechanism for segmentation, could begin to provide the
basis for an enormous number of different morphologies (Wolpert, 1989b).
THE EVOLUTION OF DIFFERENTIATION
The emergence of new cell types is a major characteristic that requires
explanation in the evolution of development. Since it is essentially present in the
cell cycle it does not, at first sight, seem to present special problems provided that
a means is available for making cells of a similar type intrinsically different, a
phenomenon we have termed non-equivalence (Lewis & Wolpert, 1976). Given
such intrinsic differences the population can diverge in evolution, making,
perhaps, one new protein at a time. In these terms the evolution of cell
differentiation is seen as adding cell types almost one at a time. A different view
has been adopted by Kauffman (1987) who sees the evolution of novel cell types
in terms of a complex combinatorial process subject to optimization. By
modelling a random ensemble for interacting genes, he shows that the system
evolves into a finite number of states which he regards as analogous to states of
cell differentiation. He has also examined the effect of selection and argues that
there is a strong tendency for trapping in local optima, “Our intuition that
selection can achieve arbitrary cell types may simply be wrong”. However, it
seems most unlikely that the state cycles in his system, are at all analogous to cell
differentiation and, more important, there is no reason to believe that evolution
acted on randomly constructed nets. Quite the contrary: evolution of cell
differentiation, as with morphology, probably proceeded by small steps. There
114
LEWIS WOLPERT
were undoubtedly constraints, but these may be of a quite different nature to
those considered by Kauffman.
Constraints on evolution due to combinatorial regulatory mechanism have
been suggested by Dickinson (1988) who has also suggested that it may account
for the diversity in patterns of gene expression among closely related species.
Attractive though this suggestion might be it is far from clear that different cell
types are characterized by a “unique combination of regulatory factors, again
drawn from a limited set”. It is worth remembering that a single gene, myo-D
can control muscle differentiation (Davis, Weintraub & Lassar, 1987) and that
single genes can switch pathways of development as with homeotic genes in
Drosophila. A criterion for combinatorial specification is that the number of genes
involved in the specification is small in relation to the total number of states. The
recent analysis of early Drosophila development by no means supports such a
conclusion. There are probably as many genes specifying segment characters as
there are segments (Ingham, 1988).
THE ORIGIN OF THE GASTRULA
The origin of the metazoa is a problem in the origin of development. Haeckel’s
views on metazoan origins have had a dominant influence. He supposed that
some protozoan became colonial and gave rise to the ‘Blastea’, a hollow sphere of
cells, the ancestor of all metazoa. Further development of the Blastea gave rise to
the ‘Gastrea’ which was a two-layered structure, the inner layer entering
through invagination at the posterior pole. A somewhat different explanation
was offered by Metschnikoff in 1886 (see Jaegerstern, 1956) and has been
promoted by Hyman (1942). Invagination is seen as a secondary process,
multiple ingression of cells being primary; in essence the idea is that phagocytosis
would be carried out by outer cells and digestion by the inner cells, while in
Haeckel’s ‘Gastrea’ the invagination provides a primitive gut with a mouth.
Thus, as Jaegerstern points out, there is much to support the view of the origin of
gastrulation as being linked to the evolution of an animal-the gastrea-whose
form corresponds to a very large number of animals a t early stages of
development. Gastrula-like forms resembling the ‘Gastrea’ can be found in phyla
as diverse as Cnidaria, Annelids, Arthropods, Echinoderms and Protochordates.
In support of the Gastrea theory Jaegerstern has provided quite a detailed
scenario for the transformation of the Blastea into a Gastrea based on its settling
on the ocean bottom and ingesting food particles from the bottom (Fig. 2). I n his
view the primitive intestine retained the cilia, and these may have assisted
feeding. He also proposes that this bottom dweller had dorso ventrality and was
bilaterally symmetrical. Bilaterality is present in the Anthozoa in the Cnidaria.
The Cnidaria provide a valuable phylum in which to try to understand the
evolution of early stages of development (Campbell, 1974). The phylum present
some of the simplest animal structures which are made up of only two germ
layers, but which develop into a variety of forms characteristic of other metazoan
phyla. There are clear signs of segmental and spacing patterns and moderately
complex spatial differentiation. Lacking a mesoblast, they are less complex
internally than other phyla.
Most Cnidaria give rise to a ciliated larva known as a planula. This is
essentially a two-layered hollow cylinder. An extraordinary feature of Cnidarian
THE EVOLUTION OF DEVELOPMENT
115
B
A
C
D
Figure 2. A possible scenario for the development of the gastrula based on Jaegerstern (1956). In A,
the ‘Blastea’may have sexual cells inside. Phagocytosis may have become limited to the ventral side
and this could have led to a small invagination to assist feeding and decomposing larger food
particles. B, Further development leads to a primitive gut with ciliary activity aiding the movement
of food particles C, D.
development is the apparent diversity of pathways from fertilization to this
rather simple structure (Fig. 3). Cleavage, which is somewhat variable
depending often on the yolkiness of the egg, is usually radial to begin with, the
first few cleavages being at right angles to one another. I n some cases, by
contrast, cleavage is anarchic. Cleavage leads to two common blastula
types-the stereoblastula which is a solid mass of cells, and the coeloblastula in
which there is a sheet of cells surrounding a hollow interior. There is, at first
sight, no obvious correlation between these two kinds of blastulae and either the
yolkiness or size of the egg (see Campbell, 1974: tableI) but it may correlate
with relative cell size (see below).
Gastrulation, the formation of a two-layered structure is no less diverse. Even
within one class there is great diversity. Gastrulation may occur by invagination
of the sheet from one pole, or by ingression of cells from one pole, or by
ingression from multiple sites, or by delamination by tangential cleavage. This
last mechanism seems to be quite different from the others, being based on a
specific plane of cleavage such that one cell remains at the embryo’s surface and
the other is effectively placed in the interior.
The other modes of gastrulation might be thought of as having a common
theme. Studies on sea urchin gastrulation have shown that cell movement and
change in cell adhesion are key processes (Gustafson & Wolpert, 1967). Initially
the mesenchyme cells enter at the vegetal pole and this is followed by
invagination of the gut: ingression and invagination are rather similar. Since
most of the different forms of gastrulation in Cnidaria involve movement of cells
from the surface layer towards the centre, the differences between them may
merely reflect spatial differences as to where this occurs and differences in the
116
LEWIS WOLPERT
adhesion of the cells to one another. Multipolar ingression is not hard to
understand, though there is the problem of specifying which cells will go in;
unipolar ingression localizes the process which involves motility and loss of
adhesion. Invagination will occur if loss of adhesion in unipolar ingression is
delayed, a feature that can be directly observed in sea urchin morphogenesis
when, in some embryos, the entry of the primary mesenchyme cells is delayed
and a small invagination occurs. It is but a small step, the retention of adhesion,
that might lead to invagination proper. So many aspects of the variability can be
seen as variations on a basic theme. But the problem then, is, why should there
be this variability? What, if any, is the selective pressure? and why should
delamination sometimes occur? But first a quite different theory about the origin
of the gastrula must be examined.
SELECTION AT THE LEVEL OF THE CELL LINEAGE
A very different interpretation of the evolution of development has been
offered by Buss (1987). Contrary to the conventional view of evolution in which
the individual is held to be the unit of selection with heritability limited to a very
small subset of genetically homogeneous cells, Buss proposes that there is selection
at the level of the cell lineage within the embryo. If, it is argued, there is
opportunity for cells that contribute to the gametes to vary, then genetic
variation arising during development “must be acknowledged as a potentially
important source of transmissible variation”. It is not easy to see what this
variation arising during development might be, more importantly, how it differs
from the variation that the germ cells can generate. Presumably both are simply
mutation. Buss argues that at the “dawn of metazoan life, the germ line was not
determined at the outset of ontogeny. Those cells which ultimately gave rise to
new individuals had to earn their position in the germ line by surviving a somatic
selective filter in the form of competitive interactions with other cell lineages
comprising the same individual”. The concept of competition between lineages
in the developing embryo is the core of Buss’s idea. Curiously, Buss believes that
there will be selection for lineages with enhanced replication whether or not the
resulting change in the lineage favours resulting the individual organism: “The
selection of lineages with enhanced replication rates proceeds independently of
whether the variant favours the individual harbouring it”. He explains the
origin of gastrulation in terms of selection for a cell lineage which abandoned
ciliation, which precluded cell multiplication, left the wall of the blastula and
entered the hollow interior. He further wishes to account for the original of other
developmental processes in similar terms “induction, competence and cell
death-are all interactions in which one cell lineage acts to limit the replication
of another while enhancing its own”.
It is not meaningful, as Buss would agree, to think of competition between cell
lineages having the same genome, for, since inheritance is controlled by that
genome no genetic distinction can be drawn between one cell lineage and
another. There is no meaningful sense in which cells with the same genetic
construction compete with one another from the point of view of evolution. The
success of each and every lineage is only dependent on the success of the
organism as a whole.
THE EVOLUTION OF DEVELOPMENT
117
.
1
A
B
C
D
Figure 3. Some different modes of development leading to the formation of the two layered planula
in Cnidaria (after Tardent, 1978). This may occur by gastrulation of the cells as A, a sheet or B, by
multipolar ingression. A quite different mode is by delamination, C. Yet a further variant starts from
a solid, rather than a hollow blastula, which then develops a cavity, D.
If a variant, a mutation, arises in development, it is now meaningful to talk
about competition between lineages. The assumption made by Buss is that if the
variant multiplies more rapidly and the germ line is not determined early, then
one of these variants will have a better chance of getting into the germ line.
Moreover, if mutants arise more frequently the more cells divide, more variants
will enter the germ line the greater the access to it is from dividing cells. So late
germ line determination could be the basis of quantitatively more mutants. But
what would happen to the rapidly multiplying variant now entering the germ
line when it develops? Buss is not explicit on this point. He makes the tacit and
untenable assumption that this cell will, when it enters the germ line, not only
generate a normal embryo, but will, moreover, now reliably produce the more
rapidly multiplying lineage a t the same stage of development. This is almost
equivalent to assuming that a malignant cell will continue to generate normal
118
LEWIS WOLPERT
cells. It is not possible to predict how a cell which undergoes a mutation during
development that increases its rate of multiplication with respect to its
neighbours will behave if it enters the germ line. There is no reason at all to
believe that its increased proliferation will be restricted to that stage in
development when the mutation occurred; it is extremely unlikely. Even more
serious an objection is that the mutant cell now generates in the best scenario an
embryo with all cells having this replicative advantage and it is hard to see how
anything but chaos will result. Moreover, since now all cells have the same
genetic constitution, and no competition is possible, the original proliferative
advantage is irrelevant and will not be selected for. Thus the whole idea of
competitive lineages as the cause of early developmental forms is based on faulty
arguments. The only way in which a mutant cell will be subject to positive
selection is if it confers advantage on the organism as a whole. Selection at the
level of the cell lineage may generate more mutants but can provide no
explanation of the forms found in early development.
SELECTION ON DEVELOPMENTAL PROCESSES
Selection on development will act to ensure reliable generation of reproductive
organisms. Reliability will be at a premium. But that selective process is not so
much on development but on the organism itself. For the evolution of
development the more interesting question is whether there is selection on the
developmental processes themselves. Is economy of energy of one developmental
process over another to be preferred? Will irrelevant development pathways be
lost? Is there a tendency to shorten a developmental process? These are not only
questions of interest in their own right, but they relate directly to the variable
pathways embryos use to achieve similar results. Are all these variants merely
drift or the results of selection for something else with a knock-on effect on
development (Sander, 1983; Dickinson, 1988).
Berrill (1949) has analysed development of scyphomedusae and claims a
correlation of egg size and type of gastrulation. I n general, he claims, the smaller
eggs gastrulate by unipolar ingression and the largest by invagination alone, the
intermediate sizes combining both methods. He rightly draws attention to the
relationship of blastula wall thickness and blastula size. The most relevant
correlation may lie here-polar ingression and invagination correlate with a thin
wall and many cells and multipolar with thick wall and few cells. Thus
differences in gastrulation may relate to the mechanics of gastrulation which are
determined by the number of cleavages prior to gastrulation. This, I would
suggest, might be related to selection for yolk/nucleus.
From his consideration of the embryology of the annelids and arthropods,
Anderson (1973) has concluded that “The blastula, or its equivalent stages of
embryonic development, has a greater stability of functional configuration than
any stage that precedes or follows it, no doubt because this is the stage at which
the fundamental framework of bodily organization is established”. Many of the
differences may be accounted for by changes in the amount of yolk and the
manner in which it is used. All insects pass through a phylotypic stage known as
the germ band (Sander, 1983), which includes head lobes, and segments of the
thorax and abdomen. Earlier stages in development may be very different, this is
in part related to yolkiness.
THE EVOLUTION OF DEVELOPMENT
119
The situation in vertebrates is similar but it is perhaps not the blastula that is
the common stage but the pharyngula-a series of paired pharynic segments,
bilateral symmetry, and a characteristic arrangement of vertebrate organ.
Ballard (1981) who coined the term, has argued that the attempt to force
homologies on earlier stages of vertebrate development is misleading. The
changes in early development in amphibians associated with the increase in yolk
have been considered by Elinson (1987) who concludes that in some large-egged
amphibians that have direct development, the embryo develops from an
embryonic disc more like that in reptiles and birds than the gastrula of other
amphibians.
Another problem arises in relation to the development of the neural tube and
somite formation. In teleost, ganoid, and cyclostomatous fish the neural plate
first forms as a thickened elongated ridge in the mid-dorsal axis, and then sinks
below the ectoderm where it develops a lumen centrally and so becomes a tube.
By contrast in the majority of chordates the outer edges of neural plate fold
inwards to form a groove and eventually a tube (De Beer, 1951). Again, in
Amphioxus, the anterior mesodermal segments initially form invaginations which
give rise to the hollow segments, whereas posteriorly the segments are a solid
mass and acquire cavities later (Balinsky, 1959). It is possible to think of these
processes as minor variations on a basic mechanism which involves cell adhesion
and polarity, but that leaves unanswered the question of preferred mechanism.
Again, it is not easy to see the similarities in the process of segmentation in
leeches, arthropods and vertebrates (see French et a/., 1988).
How important a role does energy storage and consumption play in the
evolution of development? There is the obvious correlation between the period of
development and the amount of yolk in the egg, and also the importance of
viviparity. What is much less clear, and of greater interest in the evolution of
development, is how important it is for embryos to be economical with energy. Is
energy for a developmental process an important feature? If two developmental
processes can achieve the same end result, but the one uses, say half the energy of
the other, will the more efficient process be subject to positive selection? The
answer will depend on how costly different development processes are.
In general terms our understanding of how cells use their energy is rather
crude. Studies on rabbit reticulocytes (Siems et al., 1984) which do not have a
nucleus, suggest the following major subdivisions: globin synthesis 28%, Na/K
ATPase 23%, proteolysis 15% 30% unknown. For Ehrlich ascites tumour cells
(Muller et al., 1986) the breakdown for ATP consumption is: protein synthesis
27%, proteolysis 8y0,Na/K ATPase 1 7y0,Ca2+ ATPase 1 lyo, RNA synthesis
11% and 25% unknown. These results suggest that protein synthesis and
breakdown consumes about 50% of the cell's ATP, and ion transport-as
represented by the Na/K and Ca" ATPases-about
25%. The ion transport
figure suggests that just being a functional cell and keeping the internal milieu in
order is a very expensive process. Embryos do, however, have methods for
considerably reducing energy production (Hand & Graiger, 1988) and
remaining effectively dormant for long periods, as in, for example diapause. For
sea urchins, oxygen consumption increases about five-fold after fertilization
(Yanagisawa, 1975).
Protein synthesis is clearly expensive from an energetic point of view. What
about the other developmental process such as cell movement? There is no direct
+
120
LEWIS WOLPERT
evidence but calculations on the energy requirements for neurulation or
gastrulation, based on the forces required suggest that these are very low on
energy consumption in terms of the overall production by the embryo (Gustafson
& Wolpert, 1963).
If the energy requirement for maintenance of the cell state is in fact high, as
the ion transport figures suggest, then there is a simple correlation between the
length of the developmental period and energy consumption. Rapid
development will reduce energy requirements. T h e length of the developmental
process will also correlate with the size of the animal when it begins to feed. So a
small feeding adult, or larva, is the best formula for energy conservation. If this is
so there could be selection for processes that ensure rapid development. I t is
however, important to realize that the requirement for rapid development might
be advantageous for quite different reasons, relating to the life style of the
animal. Nevertheless, selection for a particular rate of development is one of the
few selective pressures on development itself, as distinct from the phenotype to
which it gives rise.
One can thus begin to examine rapidly developing animals in order to see
what developmental modifications have occurred. To achieve rapid
development Drosophila lays down all the segments almost at the same time and
development takes about one day compared to the months for other insects. In
the hemimetabolous insects cleavage in a plectoperan may require eight days for
six synchronous divisions whereas in an orthoptera a similar number of cleavages
is achieved in nine hours (Anderson, 1972), but why should there be this
difference? Clearly there has been selection for rapid development in Drosophila
and this has required laying down in the egg large amounts of what Sander
( 1983) has called semi-finished products. Whether or not new developmental
mechanisms are involved is still not clear. This question is particularly acute in
the hymenopterans where there is total cleavage and then each cell gives rise to a
germ band and larva (Sander, 1983). It is suprising, nevertheless, that in
Drosophila the sole function of one gene nanos is to suppress the maternal product
of another gene, hunchback in the egg (Irish et al., 1989).
There are, then, enormous variations in early development and it must be
assumed that while the requirement for nutrition and environmental conditions
have provided the selective forces, the weak selection may have permitted a
variety of variants to evolve. If many of the variants are neutral this could have
important implications for the evolution of novelty, for it enables new
developmental processes to be explored and ultimately benefit the ‘adult’
phenotype.
THE CONTINUITY OF DEVELOPMENT: HETEROCHRONY AND METAMORPHOSIS
The sequence and rate of developmental processes have been a favourite
hunting ground for linking development and evolution ever since the
recapitulation theories of Haeckel (De Beer, 1951; Gould, 1977). Here we may
first note that of Von Baer’s famous laws both the first and third are just wrong:
general characters, as we have seen, do not, in early development, necessarily
appear before special characters, and neither does an animal depart more and
more, during development from the form of other animals. As we have seen, the
THE EVOLUTION OF DEVELOPMENT
121
common phylotypic stages are the blastula in many annelids and arthropods, the
pharyngula in vertebrates and the germ band in insects.
There is substantial evidence for changes in the timing of developmental
process under the general theme of heterochrony-but this is not the issue here.
The question is not what role speeding up or slowing down of a particular
process, has played, but to what extent the evolution of development has been
gradual and continuous-the continuity principle (Horder, 1983)-and whether
it has been possible to eliminate developmental stages.
For example, in insects in short germ development, most of the segments are
produced by growth and budding, but in long germ band insects like Drosophila
they arise more or less simultaneously. Superficially these two pathways seem
very different and long germ band development might be thought of as
transferring some functions from early embryogenesis back into oogenesis.
However, both might simply involve related mechanisms for specifying segments
and positional values (Sander, 1983). There is no reason to believe that any
basic process has been omitted.
The almost universal existence of phylotypic stages suggests that while
modifications are possible the changes in embryological stages are continuous.
This requires that all intermediate stages must have, at some stage been
functional or traceable to a functional stage (Horder, 1983). The origin of the
gastrula must remain our model for trying to explain the origin of other stages.
Many animals have a larval stage which may change form dramatically at
metamorphosis. This would seem to contradict the continuity requirement, for
the transition from larva to adult does not pass through stages which can be
imagined as being functional: consider for example the assembly of the imaginal
discs in insects. An explanation is that metamorphosis is always a secondarily
acquired character and the larval forms have evolved from a modification of the
original direct development, and almost buried the intermediate stages.
INDUCTION
The evolution of induction-the signalling of tissue to an adjacent tissue so as
to modify its course of development, as distinct from positional signalling
(Wolpert, 1989~)-presents a major problem. Some of the possibilities can be
illustrated with reference to the evolution of the development of the lens. As
Horder (1983) has pointed out, a plausible scenario for the development of a
light sensitive eye can be outlined. Our problem is to understand how a signal
from this eye could cause a lens to appear in the overlying ectoderm. It is in a
way a problem of spatial coordination and localization-the lens is localized in
relation to the eye. A variety of possibilities can be considered: (i) the lenses
developed in patches randomly and independently of the eye, and the eye
position was selected; (ii) lenses developed over wide areas and were stabilized
by a signal from the eye; (iii) a signal from the eye made the ectoderm above it
different from the surrounding tissue and this permitted developmental
experiments leading to a lens. In more general terms, induction can be thought
of as providing coordination between the development of different germ layers,
such as the transfer of positional information from one to the other (Wolpert,
1981).
122
LEWIS WOLPERT
REGENERATION
Regeneration is not a process that evolved separately as a useful adaptation to
damage. Regeneration is best understood as an embryonic feature persisting in
certain adults. This can be best understood in those animals that undergo
asexual reproduction by budding of fission. Taking hydra as our example,
budding can be thought of, usefully, as regeneration of a new hydra from the
body column of an adult hydra. Budding involves the complete specification and
morphogenesis of a new hydra from the adult’s body column. The processes
involved, such as specification of a new head and foot lend themselves to a
regeneration process. Thus the regeneration of a new head in hydra when the
head is cut off reflects a n organization required for asexual reproduction. Similar
considerations apply to limb regeneration in amphibia. Indeed there is direct
evidence that limb development and regeneration make use of similar processes
(Muneoka & Bryant, 1982).
CONCLUSIONS
Cells are more complicated than embryos in the sense that the intracellular
network of signals is more complex than the intercellular signals in the
developing embryo. Most of the features required for development are already
present in single cells and in single cell organisms like the Protozoa. In a curious
sense development and multicellular animals were inevitable once the cell was
invented.
A quite satisfying account of the evolution of the gastrula can be given in
terms of Haeckel’s ‘Gastrea’ theory. Given the ‘Gastrea’, then with the possibility
of segmentation and positional information a very large number of structures
become possible.
There is little reason to believe that any fundamental new developmental
principles have evolved since the early metazoa gave rise to the Gastrea. The
basic mechanisms of generating spatial differences must have been there from the
beginning. But this is not to deny that there are horrendous problems in trying to
understand the evolution of development. There are similarities but also many
differences and the latter present the most severe problems. Why are there radial
and spiral cleavage; different mechanisms for segmentation; regulative and
mosaic development? It is hard to see what selective pressures there were on
development itself, other than those related to rate and length of development. If
many variants are neutral this may be of value in the evolution of novelty.
One possible explanation for regulative development is that it is related to
asexual reproduction. This in turn provides a n explanation for regeneration
which is, in a sense, a metaphenomenon.
There are many other problems such as the evolution of novel structures such
as the neural crest, and extraembryonic membranes. All these need to satisfy a
continuity principle.
Consideration of the evolution of development may be a bit premature as we
are only just beginning to understand developmental processes. Apparently
dissimilar processes may yet reveal an underlying unity.
THE EVOLUTION O F DEVELOPMENT
I23
REFERENCES
ALBERTS, B., BRAY, D., LEWIS, J., ROBERTS, K. & WATSON, D., 1989. Molecular Biology of the Cell.
New York: Garland.
ANDERSON, D. T., 1972. The development of hemimetabolisis insects. In S. J. Counce & C. H. Waddington
(Eds), Developmental Systems: Insects: 6 1 6 3 . London: Academic Press.
ANDERSON, D. T., 1973. Embryology and Phylogeny in Annelzds and Arthropods. Oxford: Pergamon Press.
BALINSKY, B. I., 1965. A n Introduction to Embryology. Philadelphia: W. B. Saunders.
BALLARD, W. W., 1981. Morphogenetic movement and fate maps of vertebrates. American <oologist, 21:
39 1-399.
BERRILL, N. J., 1949. Developmental analysis of scyophomedusae. Biological Reviews, 24: 393410.
BODE, P. M. & BODE, H. R., 1984. Patterning in hydra. In G . Malacinski (Ed.), Pattern Formation: 213-242.
New York: Macmillan.
BUSS, L. W., 1987. The Evolution of Individuality. Princeton: Princeton University.
CAMPBELL, R. D., 1974. Cnidaria. In Reproduction of Marine Invertebrates, I : 133-199. New York: Academic
Press.
DAVIS, R. K., WEINTRAUB B. H. & LASSAR, A. B. 1987. Expression of a single transfected cDNA
converting fibroblasts to myoblasts. Cell, 51: 987-1000.
DE BEER, G . , 1951. Embryos and Ancestors. Oxford: Clarendon Press.
DICKINSON, W. J., 1988. On the architecture of regulatory systems: evolutionary insights and implications.
Bio Essays, 8 : 204-208.
ELINSON, R. P., 1987. Change in developmental patterns: embryos of amphibians with large egg. In R. A.
Raff & E. C. Raff (Eds), Development as an Evolutionary Process: 1-21. New York: A. R. Liss.
FRANKEL, J., 1984. Pattern formation in ciliated protozoa. In G . Malacinski (Ed.), Pattern Formation: 163196. New York: Macmillan.
FRENCH, V., INGHAM, P., COOKE, J. & SMITH, J., 1988. Mechanisms of segmentation. Development
(Supplement), 104: 1-251.
GOODWIN, B. C., 1989. Unicellular morphogenesis. In W. D. Stein & F. Bronner (Eds), Cell Shape:
Determination, Regulation and Regulatory Role. London: Academic Press.
GOODWIN, B. C., HOLDER, N. & WYLIE, C. C. (Eds), 1983. Development and Euolution. Cambridge:
Cambridge University Press.
GOULD, S. J., 1977. Ontogeny and Phylogeny. Harvard: Belknap.
GUSTAFSON, T . & WOLPERT, L., 1963. The cellular basis of sea urchin morphogenesis. International Review
of Cytology, 15: 139-2 13.
GUSTAFSON, T. & WOLPERT, L., 1967. Cellular movement and contact in sea urchin morphogenesis.
Biological Reviews, 42: 442498.
HAND, S. C. & GRAIGER, E., 1988. Anaerobic dormancy quantified in Artemia embryos: a calometric test of
the control of metabolism. Science, 239: 1425-1427.
HORDER, T . J., 1983. Embryological bases of evolution. In B. C. Goodwin, N. Holder & C. C. Wylie (Eds),
Development and Evolution: 315-352. Cambridge: Cambridge University Press.
HYMAN, L., 1942. The Invertebrates, Val. I I . New York: McGraw-Hill.
INGHAM, P. W., 1988. The molecular genetics of embryonic pattern formation in Drosophila. Nature, 335:
25-34.
IRISH, V., LEHMANN, R. & AKAM, M., 1989. The Drosophila posterior-group gene nanas functions by
repressing hunchback activity. Nature, 338: 646448.
JAEGERSTERN, G., 1956. The early phylogeny of the metazoa. The bilaterogastrea theory. <ool. Bidrag.
(Uppsala), 30: 32 1-354.
KAESTNER, H., 1980. Wirbellose Tiere. Jena: Fischer.
KAUFFMAN, S., 1987. Developmental logic and its evolution. Bio Essays, 6: 82-87.
LEWIS, J. H. & WOLPERT, L., 1976. The principle of non-equivalence in development. Journal of Theoretical
Biology, 62; 479490.
MULLER, M., SIEMS, W., BUTTGESIRT, F., DUMDEY, R. & RAPOPORT, S. M., 1986.
Quantification of ATP-producing and consuming processes of Ehrlich ascites tumour cells. European Journal
of Biochemistry, 161: 701-703.
MUNEOKA, K. & BRYANT, S. V., 1982. Evidence that patterning mechanisms in developing and
regenerating limbs are the same. JVature, 298: 369-37 1.
RAFF, R. A. & KAUFMAN, T . C., 1983. Embryos, Genes and Evolution. New York: Macmillan.
SANDER, K., 1983. The evolution of patterning mechanisms: gleanings from insect embryogenesis and
spermatogenesis. In B. C. Goodwin, N. Holder & C. C. Wylie (Eds), Development in Evolution: 137-159.
Cambridge: Cambridge University Press.
SIEMS, W., DUBIEL, W., DUMDEY, R., MULLER, M. & RAPOPORT, S. M., 1984. Accounting for the
ATP-consuming process in rabbit reticulocytes. European Journal of Biochemistry, 139: 101-107.
SMITH, J. M., BURIAN, R., KAUFFMAN, S., ALBERCH, P., CAMPBELL,J., GOODWIN, B., LANDE,
R., RAUP, D. & WOLPERT, L., 1985. Developmental constraints and evolution. Quarterly Review of
Biology, 60: 265-287.
124
LEWIS WOLPERT
TARDENT, P. 1978. Coelenterata, Cnidaria. In Morphogenese der Tiere. Fischer, Jena.
WEIJER, C. J., DUSCHL, G. & DAVID, C. N., 1984. Dependence of cell-type proportioning and sorting of
cell cycle phase in Dictyostelium discoideum. Journal of Cell Science, 70: 133-145.
WILCOX, M., MITCHISON, C. J. & SMITH, R. J., 1973. Pattern formation in the blue green alga,
Anabaena. I. Basic mechanisms. Journal of Cell Science, 12: 707-723.
WOLPERT, L., 1969. Positional information and the spatial pattern of cellular differentiation. Journal of
Theorefical B i o l o ~ 25:
, 1-47.
WOLPERT, L., 1981. Positional information and pattern formation. Philosophical Transactions Royal Society B,
295: 441-450.
WOLPERT, L., 1989a. Stem cells: a problem in asymmetry. In B. I. Lord (Ed.), Stem cells Journal of Cell
Science, Supplement, 10: 1-9.
WOLPERT, L., 1989b. Positional information and prepattern in the development of pattern. In A. Goldbeter
(Ed.), Theorefical Models for Cell-to Cell Signalling: London: Academic Press.
WOLPERT, L., 1989c. Positional information revisited. Development in press.
WOLPERT, L. & STEIN, W. D., 1984. Positional information and pattern formation. In G. M. Malacinski &
S. V. Bryant (Eds), Pattern Formation: 2-21. New York: Macmillan.
YANAGISAWA, T., 1975. Respiration and energy metabolism. In G. Czihak (Ed.), The Sea Urchin Embryo:
5 1&549. Berlin: Springer.