Download History and basic concepts

Historyandbasicconcepts
E
The originsof developmentalbiology
E
A conceptualtool kit
The aim of this chapter is to provide a conceptualframework for the study of
We startwith a briefhistoryof the studyof embryonicdevelopment,
development.
biologywerefirst
how someof the keyquestionsin developmental
whichillustrates
Thebig
formulated,andcontinuewith someof the essentialprinciplesof'development.
questionis how doesa singlecell-the fertilizedegg-give riseto a multicellular
into tissuesand
organism,in whicha multiplicityof differentcelltypesareorganized
body.Thisquestioncanbe studiedfrom many
organsto makeup a three-dimensional
allof whichhaveto befittedtogethertoobtaina completepicture
differentviewpoints,
andwhenandwhere;howcellscommuniof dwelopmentwhichgenesareexpressed,
fate isdetermined,howcellsproliferate
catewith eachother;howa cell'sdevelopmental
into specialized
celltypes;andhow majorchangesin bodyshapeare
anddifferentiate
is ultimatelydrivenby the
produced.We shallseethat an organism'sdevelopment
of its genes,determiningwhichproteinsarepresentin whichcells
regulatedexpression
Thegenesprovidea
andwhen,ln turn, proteinslargelydeterminehow a cellbehaves.
generativeprogramfor development,not a blueprint,as their actionsare translated
signaling,
into developmental
outcomesthroughcellularbehaviorsuchasintercellular
movement.
proliferation,
and
cell
celldifferentiation,
cell
The development of multicellular organisms from a single cell-the fertilized
egg-is a brilliant triumph of evolution. During embryonic development, the egg
divides to give rise to many millions of cells, which form structures as complex
and varied as eyes, arms, heart, and brain. This amazing achievement raises a
multitude of questions. How do the cells arising from division of the fertilized egg
become different frorn each othefl How do they become organized into structures
such as limbs and brains? What controls the behavior of individual cells so
that such highly organized patterns emerge?How are the organizing principles of
development embedded within the egg, and in particular within the genetic
material, DNA? Much of the excitement in developmental biology today comes
from our growing understanding of how genes direct these developmental
processes,and genetic control is one of the main themes of this book. Ttrousands
of genes are involved in controlling development, but we will focus only on those
that have key roles and illustrate general principles.
The development of an embryo from the fertilized egg is known as
embryogenesis.One of its first tasks is to lay down the overall body plan of
the organism, and we shall see that different organisms solve this fundamental
Ilr'
qillt
2
.
1: HtsroRYANDBAstccoNcEPTs
Fig. 1.1 Scanningelectron micrographof the
head of an aduft Drosophilomelonogoster.
Scalebar= 0.1mm.
Photogroph
byD. khorfe,fromkiencePhotoLibrory.
Fig. 1.2 Photographof a lizard,the south
easternfive-linedskink,after it hasreleased
its tail in defense.Thisspeciescandeliberately
sheditstail asa technique
to avoidcaptureby
predators;andthen regenerate
it. A pieceof
discarded
tailcanbeseenbelowthe skink.
Photogroph
Films.
fromOxfordScientiftc
problem in severalways. The focus of this book is mainly on animal developmentthat of vertebrates such as frogs, birds, fish, and mammals, and of a selection of
invertebrates, such as the sea urchin, ascidians, and, above all, the fruit fly
Drosoplnlamelarwgaster
(Fig. 1.1) and the nematode woran Caenorhabditis
elegans.ltis
in these last two organisms that our understanding of the genetic control of
development is most advanced, and the main features of their early development
are considered in Chapters 2 and 5, respectively. In Chapter 6, we look briefly at
some aspects of plant development, which differs in some respects from that of
animals but involves similar principles.
The development of individual organs such as the vertebrate limb, the insect eye,
and the nervous system illustrates multicellular organLation and tissue differentiation at later stagesin embryogenesis,andwe consider someof these systemsin deail
in Chapters 8-10. We also deal with the development of sexual draracteristics
(Chapter 11). The study of developmental biology, however, goes well beyond
the development of the embryo. We also need to understand how some animals can
regenerate lost organs (Fig. 1.2, and Chapter 13), and how post-embryonic growth
of the organism is contrclled, a process that includes metamorphosis and aging
(Chapter 12). Taking a wider view, we consider in Chapter 14 how developmenal
mechanismshave evolvedand how they constrain the very processof evolution itself,
One might ask whether it is necessary to cover different organisms and
developmental systems in order to understand the basic features of development.
The answer at present is yes. Developmental biologists do indeed believe that there
are general principles of development that apply to all animals, but that life is too
wonderfi,rlly diverse to find all the answers in a single organism. As it is, develop
mental biologists have tended to focus their efforts on a relatively small number of
animals, chosen originally becausethey were convenient to study and amenable to
experimental manipulation or genetic analysis.This is why some creatures, such as
the frog Xenopu laais (Fig. 1.3), the nematode Caenorlwbditis,and the fruit fly
Drosophila,have such a dominant place in developmental biology, and are encountered again and again in this book. Indeed, it is very encouraging that so few
systemsneed to be studied in order to understand animal development. Similarly,
the thale-cress,Arabiilopsisthahana,can be used as a model plant to consider the
basic features of plant development.
One of the most exciting and satis$ring aspectsof developmental biology is that
understanding a developmental process in one organism can help to illuminate
similar processeselsewhere,for example in organismsmuch more like ourselves.
Nothing illustrates this more dramatically than the influence tltat our understanding
of Drosophiladevelopment, and especially of its genetic basis, has had throughout
developmental biology. In particular, the identification of genes controlling early
embryogenesis in Drosophilahas led to the discovery of related genes being used
in similar ways in the development of mammals and other vertebrates. Such discoveries encourageus to believe in the emergenceof general developmental principles.
Frogs have long been a favorite organism for studying development because
their eggsare large, and their embryos are robust, easyto grow in a simple culture
medium, and relatively easy to experiment on. The South African frog Xenop.tsis
the model organism for many aspects of vertebrate development, and the main
features of its development (Box 1A, pp. 4-5) serve to illustrate some of the basic
stages of development in all animals. The early development of Xenopw and the
other model vertebratesis discussedin Chapters3 and 4.
In the rest of this chapter we first look at the history of embryology-as the
study of developmental biology has been called for most of its existence. The term
BL
IOLOC Y
TH E OR IC IN SOF D E V E LOP M E N TA
f,l of
ffy
ti s
io f
tnt
rat
to f
EF,
developrnental biology itself is of much more recent origin. We then introduce
somekey conceptsthat are used repeatedly in studying and understanding develop
ment.
biology
Theoriginsof developmental
Many questions in embryology were first posed hundreds, and in some cases
thousands,ofyears ago.Appreciating the history ofthese ideas helps us to understand why we approach developmental problems in the way that we do today.
fi&
rail
tics
md
GII
rrh
rnc
rrtal
relf.
md
ENL
tere
itoo
lop
lo f
hto
ta s
tf ly
nrnfew
lrly'
: the
tbat
nate
lves.
di:rg
hout
Early
used
tcovples.
zuse
lure
rls is
nain
besic
il the
r tle
tern
r.r Aristotle first defined the problem of epigenesisand preformation
A scientific approach to explaining development started wittt HipPocratesin Greece
in the Sth century nc. Using the ideas current at the time, he tried to explain
developmentin terms of the principles of heat, wetness,and solidification. About
a century later the study of embryology advanced when the Greek philosopher
Aristotle formulated a question that was to dominate much thinking about development until the end of the 19th century.Aristotle addressedthe problem of how
the different parts of the embryo were formed. He considered two possibilities: one
was that everything in the embryo was preformed from the very beginning and
5imply got bigger during development; the other was that new structures arose
progressively,a processhe termed epigenesis(which means 'uPon formation') and
that he likened metaphorically to the 'knitting of a net'. Aristotle favored epigenesis
and his conjecturewas correct.
Aristotle's influence on Europeanthought was enounous and his ideasremained
dominant well into the 17th century. The conffary view to epigenesis,namely that
the embryo was preformed from the beginning, was championed anew in the late
17th century. Many could not believe that physical or chemical forces could mold
a living entity like the embryo. Along with the contemPoraneous background of
belief in the divine creation of the world and all living things, was the belief that
all embryos had existed from the beginning of the world, and that the first embryo
of a speciesmust contain all future embryos.
Even the brilliant 17th century Italian embryologist Marcello Malpighi could not
free himself from preformationist ideas.While he provided a remarkably accurate
description of the development of the chick embryo, he remained convinced,
against the evidence of his own observations, that the embryo was already present
from the verybeginning (Fig.1.4).He arguedthat atvery early stagesthe parts were
so small that they could not be observed,even with his best microscope.Yet other
preformationists believed that the sperm contained the embryo and some even
claimed to be able to see a tiny human-an homunculus-in the head of each
human sperm (Fig.1.5).
The preformation/epigenesis issue was the subject ofvigorous debate throughout
the 18th century. But the problem could not be resolved until one of the great
advancesin biology had taken place-the recognition that living things, including
embryos,were composedof cells.
made
Malpighi's
drawings,
Thefigureshows
of thechickembryo.
description
Fig.1.4 Malpighi's
(bottom).
Hisdrawings
accurately
(top),andat2 days'
incubation
theearlyembryo
in1673,
depicting
oftheembryo.
andbloodsupply
illustrate
theshape
Reprintedby permissionof the Presidentond Councilof the RoyolSociety.
E
Fig. 1.3 Photographofthe adult South
Africancfaw-toedfrog, )knopus loevis.
Scalebar= 1 cm.
courtesy
ofl. Smith.
Photogroph
3
4
t: Htsro RyANDBAs t cc oNc Epr s
BoxlA
Basicstages
of Xenopus
|oevisdevetopment
@&
Egg
*ffir{eee
%u:*
@
(section)
Gastrula
Vegeratpole
/
/
("/
^&
Fettilization
Gastrulation
Metamorphosis
Free-swimming
tadpole
ftrture
| 5ta 9e12I
Al
(section)
venrlar
t(.
0rganogenesis
\\
\\
Antedor
\\==-
'*--4ffii
;'J'#ff/
"""'
ll&i{
somiteslffii'|j
,f^l
/
"' 1
u* un)
iooo.r
{w t\l
-\p
rorkmass
\-
Posteriol
Tailbud-sta9e
embrlo
(dorsalviewwith
surfareremoved)
==-f
/
Tailbud*tage
embryo
(latentiiew)
Antedor
ffi
neural
folds
Neurula
I Stage13|
(surface
removed
to reveal
notochord)
Postelior
@
(dorsal
view)
@
C oni --.
IH E O R IGIN S O F D EVELOPM ENTA L B I O L O G Y
Box 1A (continuedl
Basic stages of Xenopusloevisdevelopment
Although vertebrate development is very varied, there are a
number of basic stages that can be illustrated by following
the developmentof the frog Xenopusloevis,which is a favorite
ectoderm, which forms both the epidermisand the nervous
system. The future endoderm and mesoderm, which are
destinedto form internalorgans,are still on the surfaceof the
organism for experimentalembryology.The unfertilizedegg
is a largecell.lt hasa pigmenteduppersurface(the animal pole)
embryo. During the next stage-gastrulation-there is a
dramatic rearrangementof cells;the endodermand mesoderm
move inside,and the basicbody plan of the tadpole is estab-
and a lower region (the vegetal pole) characterizedby an
of yolkgranules.Soevenat the beginning,the egg is
accumulation
cellsfrom the animalhalf
development,
not uniform;in subsequent
becomethe anterior(head)endof the embryo.
After fertilizationof the egg by a sperm,and the fusionof male
and female nuclei,cleavage begins.Cleavagesare mitotic divisionsin which cellsdo not grow betweeneachdivision,and so
cleavagesthe cellsbecomesmaller.After about
with successive
12 divisioncycles,the embryo,now knownasa blastula,consists
of many small cellssurroundinga fluid-filledcavity (the blastocoel) abovethe largeryolky cells.Already,changeshaveoccurred
within the cellsand they haveinteractedwith eachother so that
some future tissue types-the germ layers-have become
partly specified.The future mesoderm for example,which gives
riseto muscle,cartilage,bone,and otherinternalorganslikethe
heart,blood,and kidney,is presentin the blastulaasan equatorial band.Adjacentto it is the future endoderm, which givesrise
to the gut, lungs,and liver.The animalregionwill give riseto
lished.Internally,the mesodermgivesriseto a rod-likestructure
(the notochord), which runs from the headto the tail, and lies
centrallybeneaththe future nervoussystem.On either side of
the notochord are segmented blocks of mesoderm called
somites, which will give rise to the musclesand vertebral
column.aswellasthe dermisof the skin.
Shortly after gastrulation,the ectoderm above the notochord
foldsto form a tube (the neural tube), which givesriseto the brain
and spinalcord-a processknown as neurulation. By this time,
other organs,suchas limbs,eyes,and gills,are specifiedat their
future locations,but onlydevelopa little later,duringorganogenesis.
cellssuch as muscle,cartilage,
specialized
Duringorganogenesis,
Within48 hoursthe embryohasbecome
and neuronsdifferentiate.
a feeding tadpole with typical vertebrate features. Becausethe
conditiming of eachstagecanvary,dependingon environmental
andotherembryosareoften
stagesin Xenopus
tions,developmental
denotedby stagenumbersratherthan by hoursof development.
r.z Celltheory changedthe conceptionof embryonic development
and heredity
The cell theory developedbetween 1820 and 1880 by' among others, the German
botanist Matthias Schleidenand the physiologistTheodor Schwann,was one of the
most illuminating advancesin biology, and had an enorrnousimpact. It was at last
recognizedthat all living organisms consist of cells, which are the basic units of
life, and which ariseonly by division fiom other cells.Multicellular organismssuch
as animals and plants could now be viewed as communities of cells. Development
could not therefore be based on preformation but must be epigenetic, because
during development many new cells are generatedby division fiom the egg, and
new tj4)es of cells are formed. A crucial step forward in understanding development was the recognition in the 1840sthat the egg itself is but a single, albeit
cell.
specialized,
An important advancewas the proposal by the 19th century German biologist
August Weismann that the offspring does not inherit its characteristicsfrom the
body (the soma)of the parent but only from the germ cells-egg and sperm-and
that the germ cells are not influenced by the body that bears them. weismann
thus drew a fundamental distinction between germ cells and body cells or
wascurledupin theheadof each
thatanhomunculus
believed
Fig.1.5 Somepreformationists
sPerm,
(1694).
Hortsoeker
An imoginotivedrowing,ofter Nicholos
r
6
l: HtsroRyANDBAstccoNcEprs
| | mutationsin
||;:T;#:.f1t,f,,,.
/a<l
Et
r
E ll
Fig. 1.6 Thedistinction between germ cells
and somaticcells.In eachgenerationgermcells
giveriseto both somaticcellsandgermcells,
but inheritance
isthroughthe germcellsonly.
Changes
that occurdueto mutationin somatic
cellscanbe passed
onto theirdaughtercellsbut
do not affectthe germline.
Fll
ll
ll
o-ol{{
zlgote .+
germcel l s
j- o- , =
/
n . . lE
/ /t' l a
/d € < )
o -:
_ o-o-|t{=o-oll{
L-------___a__| | zygote
germcel l s
ll
| | zygote L_______________
germcel k
l l '-
somatic cells(Fig.1.6).characteristicsacquired by the body during an animal's life
cannot be transmitted to the germline. As far as heredity is concerned, the body is
merely a carrier of germ cells. As the English novelist and essayist samuel Butler
put ic "A hen is only an egg's way of making another egg.,'
work on sea urchin eggs showed that after fertilization the egg contains two
nuclei, which eventually fuse; one of these nuclei belongs to the egg while the
other comes from the sperm. Fertilization therefore results in an egg carrying
a nucleus with contributions from both parents, and it was concluded that the cell
nucleus must contain the physical basis of heredity. The climax of this line of
research was the eventual demonstration, toward the end of the 19th century,
that the chromosomes within the nucleus of the fertilized egg-the zygote-are
derived in equal numbers from the two parental nuclei, and the recognition that
this provided a physical basis for the transmission ofgenetic characters according
to laws developedby the Austrian botanist and monk Gregor Mendel. The constancy
of chrornosome number from generation to generation in the somatic cells was
found to be maintained by a reduction division (meiosis) that halved the chromosome number in the germ cells, whereas somatic cells divide by the processof
mitosis, which maintains chromosome number. The precursors to the germ cells
contain two copies of each chromosome, one maternal and one paternal, and are
called diploid. This number is halved by meiosis during formation of the gametes,
so that each germ cell contains only one copy of each chromosome and are called
haploid. The diploid number is restored at fertilization.
r.3 Two main types of development were originally proposed
once it was recognized that the cells of the embryo arose by cell division from the
zygote, the question emerged of how cells becarne different from one another.
with the increasing emphasison the role of the nucleus, in the 1gg0sweismann
put forward a model of development in which the nucleus of the zygote contained
a number of specialfactors or determinants (Fig. 7.7).He proposedthat while the
fertilized egg underwent the rapid cycles ofcell division known as cleavage,these
determinants would be distributed unequally to the daughter cells and so would
control the cells' future development. The fate of each cell was therefore predetermined in the egg by the factors it would receive during cleavage. This type
of model was termed fmosaic', as the egg could be consideredto be a mosaic of
B IIOLOGY
TH E OR IC IN SOF D E V E LOP MEN TA
I-l
cl
"l
Flg. 1.7 Weismann'stheory of nuclear
determination.Weismannassumedthat there
werefactorsin the nucleusthat weredistributed
to daughtercellsduring
asymmetrically
cleavageanddirectedtheirfuturedevelopment.
EI
I
,l
__t
life
ly is
der
trro
the
Fng
:cell
eof
Ery'
-are
that
ilirg
|ncy
was
lne
rs of
rells
il are
ttes,
rlled
rttre
ther.
n;rnn
dned
e the
lhese
Fuld
rede
i tyPe
rk of
rr
7
discretelocalized determinants.Central to Weismann'stheory was the assumption
tbat early cell divisions must make the daughter cells quite different from each
other as a result ofunequal distribution ofnuclear components.
In the late 1880s,initial support for Weismann's ideas came from experiments
carried out independently by the German embryologist Wilhelm Roux, who experimented with frog embryos. Having allowed the first cleavage of a fertilized frog
eg, Roux destroyed one of the two cells with a hot needle and found that the
remaining cell developed into a well-formed half-larva (Fig. 1.8).He concluded that
the 'development of the frog is based on a mosaic mechanism, the cells having
rheir character and fate determined at each cleavage".
But when Roux's fellow countqrman, Hans Driesch,repeatedthe experiment on
seaurchin eggs,he obtained quite a different result (Fig.1.9).He wrote later:
'But things turned out as they were bound to do and not as I expected; there was,
typically, a whole gastrula on my dish the next morning, differing only by its small
size from a normal one; and this small but whole gastrula developed into a whole
md typical lawa."
Driesch had completely separated the cells at the two-cell stage and obtained a
normal but small larva. That was just the opposite of Roux's result, and was the
first clear demonstration of the developmentalprocessknown as regulation. This
is the ability of the embryo to develop normally even when some portions are
renoved or rearranged. We shall see many examples of regulation throughout
the book (an explanation of Roux's experiment, and why he got the result he did,
is also given later, in Section3.7).
r.l The discoveryof induction showed that one group of cellscould determine
the developmentof neighboring cells
Ahhough the concept of regulation implied that cells must interact with each other,
fre central importance of cell-cell interactions in embryonic development was not
really establisheduntil the discovery of the phenomenon of induction, in which
me cell, or tissue, directs the development of another, neighboring, cell or tissue.
The importance of induction and other cell-cell interactions in developmentwas
pmved dramatically in7924when Hans Spemannand his assistantHilde Mangold
wried out a famous transplant experiment in amphibian ernbryos. They showed
&at a partial second embryo could be induced by grafting one small region of
m early newt embryo onto another at the same stage (Fig. 1.10).The grafted
tissue was taken from the dorsal lip of the blastopore-the slit-like invagination
frat forms where gastrulation begins on the dorsal surface of the amphibian
rmbryo (seeBox 1,4.,pp. 4-5). This small region they called the organizer, since it
seemedto be ultimately responsible for controlling the organization of a complete
;lfr
8
t: HtsroRYANDBAstccoNcEPTs
Fig. 1.8 Roux'sexperimentto investigate
Weismann'stheory of mosaicdevelopment.
Afterthe first cleavageof a fiog embryo,oneof
thetwo cellsiskilledby prickingit with a hot
needle;
the otherremains
undamaged.
At the
blastula
stagethe undamaged
cellcanbeseen
to havedividedasnormalintomanycellsthatfill
halfoftheembryo.Thedevelopment
ofthe
blastocoel
is alsorestrictedto the undamaged
half.Inthe damagedhalfof the embryo,no cells
appearto haveformed.At the neurulastage,
the undamaged
cellhasdeveloped
into
something
resembling
halfa normalembryo.
Fig. 1.9 Theoutcomeof Driesch's
experimenton seaurchin embryos,which
first demonstratedthe phenomenonof
regulation.Afterseparationof cellsat the twocellstage,the remainingcelldevelopsinto a
small,butwhole,normallarva.Thiscontradicts
Roux's
earlierfindingthat if oneofthe cellsof a
two-cellfrogembryoisdamaged,
the remaining
celldevelops
intoa half-embryo
only(seeFig.1.8).
embryonic body, and it is commonly known as the Spemann-Mangoldorganizer,
or just the Spemannorganizer. For their discovery,Spemannreceived the Nobel
Prize for Physiology or Medicine in 1935, one of only two ever given for embryological research. Sadly, Hilde Mangold had died earlier, in an accident, and so could
not be honored.
1.5 The study of developmentwas stimulated by the coming together
of geneticsand development
During much of the early part of the twentieth century there was little connection
between embryolory and genetics.When Mendel'slaws were rediscoveredin 1900
there was a great surge of interest in mechanisms of inheritance, particularly in
relation to evolution, but less so in relation to development.Geneticswas seenas
the study of the transmission of hereditary elements from generation to generation, whereas embryolory was the study of how an individual organism develops
and, in particular, how cells in the early embryo became different from each other.
Geneticsseemed,in this respect,to be irrelevant to development.
BL
IOLOGY
TH E OR IGIN SOF D E V E LOP ME NTA
graftedfinman
lip of blastopore
andMangoldof inductionof a newmain
Dorsal
bySpemann
Fig.1.10 Thedramaticdemonstration
of newtto the
spedes
from
unoicrirented
A pieceoftissue(yellow)
gastrula.
amphibian
eariy
in
the
region
spe<ies
bodyaxisbytheorganizer
blasioioelroofof a pigmented
a
of
side
opposite
to
the
grafted
is
gastrula
a newt(Iritoncristotus)
thedorsallipoftheblastopore-of
gastrufafromanother,pigmented,newtspecies(Tritontoeniotus'pink)'Thegraftedtissueinducesa
newbodyaxiscontainingneura|tubeandsomites.Theunpigmentedgrafttissueformsanotochord
atitsnewsite(seesectionin|owerpane|)buttheneura|tubeandtheotherstructuresofthenewaxis
and
byspemann
discovered
region
Theorganizer
hosttissue.
fromthepigmented
havebeeninduced
organizer'
isknownastheSpemann
Mangold
to link genetics and embryoloey
An important concePt'that eventually helped
phenotype. This was first put forward
was the distinction between genotype and
in 1909' The genetic endowment of an
by the Danish botanist Wilh"l* ;oh"tt"'"n
r,
EI
lgi
ild
rcn
900
7in
Das
Eralops
fur.
organism_thegeneticinformationitacquiresfromitsparents_isthegenotype'
ItsvisibleaPpearance,internalstructufe,andbiochemistD/atanystageofdevelopmentisthephenotyPe.WhilethegenotypecertainlycontrolsdeveloPment'envigenotype influence the phenotype' Despite
ronmental factors interacting with the
considerable differences in
identical genogpes, identical twins can develop
having-phenotypes
tend to become more
as ttrey grow up (Fig' 1'11)' and these
ttreir
posed in terms of the
be
now
could
evident with age.The problem of development
genetic endowment
the
how
relationship between genotyPe and phenotype;
give
rise to a functioning
to
during development
becomes'ffanslated' o,;"*p'"""a'
organism.
was a slow and tortuous
The coming together of genetics and embryology
and function of the genes
process. Little progress was made until the nature
in the 1940s that genes encode
were much better understood' The discovery
proteinswasamajorflmingpoint.Asitwasalreadyclealthatthepropertiesofa
the fundamental role of genes in
cell are determined Uy tne f,roteins it contains'
developmentcouldatlastbeappreciated.BycontrollingwhichProteinswere
in cell proPerties and behavior
made in a ceII, genes could control the changes
(induced)
embrYo
setondary
thatoccurredduringdevelopment.Afurthermajoradvanceinthelg60swasthe
understandingthatsomeg"n"s",,cod"proteinsthatcontroltheactivityofother
genes.
r--l
rr
€=
I
FI
F
E
r.6Deve|opmentisstudiedmain|ythroughaselectionofmode|organisms
Althoughthedevelopmentofawidevarietyofspecieshasbeensttrdiedatonetime
oranother,arelativelysmallnumberoforganismsprovidemostofourknowledge
thus regard them as models for
about developmental mechanisms' We can
understandingtheprocessesinvolvedandtheyareoftencalledmode|organisms.
Seaurchinsandamphibianswerethemainanimalsusedforthefirstexperimentalinvestigationsatthebeginningofthetwentiethcenturybecausetheirdevelop.
large
obtain and' in the caseof the frog' sufficiently
ing embryos are both
""'y1o
androbustforrelatively"",y"*p",i*entalmanipulation,evenatquitelatestages'
(Ws mtsathrs)'the chicken (Gallus
Among vertebrates, th; fro; X'bevis' the mouse
gollus),andthezebrafish(Dantoreriol'arethemainmodelorganismsnowstudied'
.{monginvertebrates,thefruitflyD.me|anogasterandthenematodeworrnC.e|egans
a great deal is known about their
have been the focus ofmost attention, because
easily genetically modified' With the
developmental genetics and they can also be
adventofmodernmethodsofgeneticanalysis,therehasalsobeenaresurgenceof
For plant developmental
interest in the sea urchin stronglocentrotuswluratus.
organism' The life
model
main
the
as
hiology, the thale-cress L thakann serves
Fig. 1.11 Thedifferencebetween genotype
and phenotype.Theseidenticaltwins havethe
onefertilizedeggsplit
samegenotypebecause
Theirslight
development'
during
two
into
isdueto nongenetic
in appearance
difference
influences'
factors,suchasenvironmental
of JosiondloimePoscuol'
courtesy
Photogroph
10
A N D B A 5 ICCONCEPT 5
1: H I S T O R Y
cyclesandbackgtounddetailsforthesemodelorganismsaregivenintherelevant
chapters later in the book.
firereasonsforthesechoicesarepartlyhistorical-onceacertainamountof
Iesearchhasbeendoneononeanimalitismoreefficienttocontinuetostttdy
with another species-and partly a
it rather than start at the beginning again
Each specieshas its advantages
question of ease of study and biotogicat i"t:tIt
anddisadvantagesas"aet'etopment"tmodel'Thechickembryo'forexample'has
eggsare
of vertebrate developmentbecause fertile
longbeen studied as *
"*"*ft"
experimental microsurgical manipulation
easily available, the embryo
-i'tt"*a'
egg' A disadvantage' however' was
very well, and it can be cultured outside the
genetics.Howevet we know
that little was known about the chick's devel0pmental
the mouse is more difficult
although
about the genetics of the mouse'
;;;;";
tostudyinsomewaysasdevelopmenttakesplaceentirelywithinthemother.
Manydevelopmentalmutationshavebeenidentifiedintfiemouse,anditisalso
techniques' It is also the best
amenable to genetic modification by transgenic
mammalian development' including
experimental model we have for studying
th a to fh u n ans.Thezebrafi shi samorerecentaddi ti ontothesel ectl i stof are
in large numbers' the embryos
vertebrate model systems; it is easy to breed
transparentandsocelldivisionsandtissuemovementscanbefollowedvisually,
and it has great potential for genetic investigations'
to understand how genes control
A major goal of developniental biology is
embryonicdevelopment,andtodothisonemustfirstidentiffthosegenescritiThis task can be approached in a variety
cally involved in controlling development'
ofways,dependingonthe"organisminvolved,butthegeneralstaltingpointistlre
identificationofmutationsthatalterdevelopmentinsomespecificandinformativeway,asdescribedinthefollowingsection;Techniquesforidentiffinggenes
manipulating their expression in the
that control development and detecting and
along with techniques for genetic
organism are described throughout the book'
maniPulation.
S o m e o f ourmodel organi smsaremoreamenabl etoconventi onal genet
ic
in developmental biology' little
analysis than others. Despite its importance
has the disadvantage of
conventional genetics has teen done on Xlnwis'which
atetrap|oidgenome(onewithfoursetsofchromosomesinthesomaticcells)and
years to reach sexual maturity. Using
a relatively long breeding period, taking 1-2
however' many developmental genes
modern genetic ana uioiiformatics methods'
sequencecomparison with known
have been identifled ir_X'lastisby direct DNA
genesinorganismssuchasDrosoptn|nandmice.ThecloselyrelatedfrogXenoptls
(Silurana)troptcaltsisamuchmoreattractiveorganismforgeneticanalysis;itis
to produce ffansgenic organisms'
diploid and can also be genetically manipulated
identify develop
is now being sequenced'which will help
T:heX. troprtcalisgenome
mentalgenesinX.laevis.Thesituationwithbirdshasbeenrathersimilar'buthere
can be used to identiff develalso, direct DNA comparisons with other organisms
the chicken genome has now been
oPmental g"n"s. tt e DNA sequence of
completed,whichshouldbeaconsiderablehelpinthegeneticanalysisofdevelop
mentinthisspecies.Sequencingofthegenomeoft}reseaurchinS'plpwaws,
biology, is also under way'
one of the oldest model organirms i" developmental
Completegenomer"qo""t"'ht""beenavailableforsomeyearsforhuman'mouse'
Drosoplila, and Caensrlwb ditis'
gene tral been identified in
In general, when an important developmental
consider whether a corresponding
one animal, it has proved iery ruw"rdirrg to
BL
IOLOGY
TH EOR IGIN SOF D E V E LOP ME N TA
EC
gene is present and is acting in a developmental capacity in other animals.
Such genes are often identified by a sufficient degree of nucleotide sequencesimilarity to indicate descent from a conrmon ancestral gene. Genes that meet this
criterion are known as homologous genes. As we shall see in Chapter 4, this
approach identified a hitherto unsusPected class ofvertebrate genes that speciff
the regular segmented pattern from head to tail that is represented by the different grpes of vertebrae at different positions. These genes were identified by
their homolory with genes that speciff the identities of the different body
DII
segments in Drosoplfila.
trt
of
ty
'a
Fs
Ets
11
a5
tr/
Elt
Ef.
bo
ESt
ng
of
EC
[y,
rol
itiEty
he
nirDes
fhe
ctic
rtic
ttle
:o f
nd
ing
nes
rvn
|p{ls
r.z The first developmentalgeneswere identified as sPontaneousmutations
Most of the organisms dealt with in this book are sexually reproducing diploids:
their somatic cells contain two copies of each gene, with the exception of those on
sex chromosomes.X. lawis is an exception, as it is tetraploid; that is' it has four
copies of the basic genome in its somatic cells, and two copies in the germ cells.
This means that it will have up to four copies of each gene in its somatic cells,
which complicatesgenetic analysis.In a diploid sPecies,one copy,or allele, of each
gene is contributed by the male parent and the other by the female. For many
genes there are several different 'normal' alleles present in the poPulation,
which leads to the variation in normal phenotype one seeswithin any sexually
reproducing species.Occasionally,however, a mutation will occur spontaneously
in a gene and there will be marked change, usually deleterious, in the phenotype
of the organism.
Many ofthe genes that affect development have been identified by spontaneous
mutations that disrupt their function and Produce an abnormal phenotype.
Mutations are classifiedbroadly according to whether they are dominant or recessive
(Fig. 1.12). Dominant and semi-dominant mutations are those that produce
a distinctive phenotype when the mutation is present in just one of the alleles of
mutation(e.9.wfiite-)
Recesive
Genotype
Phedotype
Genotype
Pheno0pe
wildtype
white+
normal
wildtype
nomal
tis
w{K
llrs.
j@x.
bq
white+
Ere
:ve1-
fiIs,
r:ry.
nse,
r
nomal
\
.e,X;ffi;
@K
white-
mutation
homozygous
whitt
ili n
iling
re
\,
mutation
heterozygous
Ben
bp
mutation(e.9.Bnchryryl
Semi-dominant
x!,
x@{.
whitt
ir
re
f/\
eyes
white
@
+
X}@(K
i*@&
+
mutation
heterozygous
e@e
r"sGWX
"d{-\
'*S*+*-t
,....*,***,1,
tail
deformed
\*fu+s
",il*^\
T
mutation
homozygous
T
xs@{K
X@K
T
lethal
embryonic
{,&
Fig.1.12 Typesof mutations.Left:a mutation
whenit onlyhasaneffectin the
is recessive
state;that is,whenboth copiesof
homozygous
the genecarrythe mutation.Right:by contrast,
mutation
a dominantor semi-dominant
in the
produces
aneffecton the phenotype
state;that is,whenjust onecopy
heterozygous
A plussign
of the mutantgeneispresent.
denoteswildtype,anda minussignrecessive.
f isthe mutantform of the geneBrochyury'
t2
1 : HIS T ORY A N D BASIC CONCEPTS
a pair; that is, it exerts an effect in the heterorygous state. Dominant mutations
can be missed as they are often lethal. By contrast, recessive mutations such as
white in Drosopli)a,which produces flies with white eyes rather than the normal
red, alter the phenotype only when both alleles of a pair carry the mutation; that
is, when they are homozygous.
In general, dominant mutations are more easily recognized, particularly if they
affect gross anatomy or coloration, provided that they do not causethe early deati
of the embryo in the heterozygous state. Truly dominant mutations are rare,
however. A mutation in the mouse gene Braclryuryis a classic example of a semidominant mutation and was originally identified because mice heterozygous for
this mutation (symbolized by Q have short tails. When the mutation is homozygous it has a much greater effect, and embryos die at an early stage, indicating that
the gene is required for normal embryonic development (Fig.1.13).Oncebreeding
studies had confirmed that the Brachyny ffait was due to a single gene, the gene
could be mapped to a location on a particular chromosome by classical genetic
mapping techniques. IdentiSring recessive mutations is more laborious, as the
heterozygote has a phenotype identical to a normal wild-type animal, and a carefully
worked-out breeding program is required to obtain homozygotes. Identiffing
potentially lethal recessivedevelopmental mutations requires careful observation
and analysis in mammals, asthe homozygotesmaydie unnoticed inside the mother.
Very rigorous criteria must be applied to identit/ those mutations that are affecting a genuine developmental process and not just affecting some vital but routine
housekeeping function without which the animal cannot suryive. One simple
criterion for a developmental mutation is embryonic lethality, but this also catches
mutations in genes involved in housekeeping functions. Mutations that produce
abnormal patterns of embryonic development are much more promising candidates
for true developmental mutations. In later chapters we shall see how large-scale
screening for mutations following mutagenesis using chemicals or X-rays has
identified many more developmental genes than would ever have been picked up
by rare spontaneous mutations.
& o @.
-:.=*.-/.,
Shorttail
;1
Fig. 1.13 Geneticsof the semi-dominant
mutation Brcchyury(I)in the mouse.A male
heterozygote
carryingthe f mutationmerely
hasa shorttail.Whenmatedwith a normal
(wildtype,++)female,someof the offspringwill
alsobe heterozygotes
andhaveshorttails.
Matingtwo heterozygotes
togetherwill resultin
someofthe offspringbeinghomozygous
(I/f
for the mutation,resulting
in a severe
andlethal
developmental
abnormality
in whichthe
posteriormesodermdoesnot develop.
r
H@,@dre\
=-'5)v,.-:+*ffi
iffi@@
ffiifiHy,ii;iits.,
K IT
A C ON C E P TUATOOL
L
I
5
I
t
v
h
Ii
5ummary
startedwith the Greeksmorethan 2000years
Thestudvof embryonicdevelopment
not containedcomp|ete|y
,oo' n'i'totru put forwardthe ideathat embryoswere
emergedgradually'
structure
and
form
that
but
egg,
,ne
*',n,^
;eformed"in
by
"i;.irr"
in the 17thand18thcenturies
Thisideawaschallenged
developed.
.ri6u
orever
"tUruo
had
been,
that
embryos
at
thit
idea
il;;ijJ,"*i',rior"t"rn.'roon,,r,"
of the cell
The emergence
*""fa i", had existedfrom the beginningof the world'
t heor v int h e l g th c e n tu ry fi n a | | y s e tt| e dthei ssuei nfavorofepi genesi s' andi tw as
someofthe
cells'
specialized,
hishly
albeit
lss;'u
''"gr",
to regutate,
able
are
embryos
urchin
,io*"4 iirt u"ryu.1ysea
the
established
rhis
killed'
or
it."ttr.r"
'"mou.d
"u"n
communication
on
in
least
at
mustdepend
Part
tiril"u"ropment
rg
:";rJilffi;;;;
.#;;;;*"-r,t,
il;;il;;;;;iv
;;;;il;..il"
[e
#;";;
dc
fi",#;;;il;;;;;;."r*
;;f
;;ilil;;;;
}I
vet
lle
[y
ng
on
er.
of cell-cell
6," ."it, of the embryo.Direct evidencefor the importance
and
outbvspemann
carried
n'.n experiment
could
resion
orsanizer
,r,-.,,r,".-"rr,or ihe amphibian
embrvo'
intoanother
whentransplanted
;;;;;;;;
;tiai"murv-otromhosttissue
inthe
hasonlybeenfullyappreciated
development
genes
incontrolling
ii" J"
much
made
"rir,i.iJii"'rr"lr
hasbeen
of development
the-seneticbasis
;;;,;;;
"ftechniques
biology'
of molecular
timesbythe
in recent
easier
rrbe
Ple
[es
tlce
tres
ale
bas
lu p
A conceptualtoolkit
complicated fate a single
Development into a multicellular organism is the most
and the challenge of
living cell can undergo; in this lies both the fascination
developmentalbiology.Yetonlyafewbasicprinciplesareneededtostafttomake
,"rrr"ofdevelopmentalprocesses.Therestofthischapterisdevotedtointroducrepeatedly throughout
ing these key concepts.These principles afe encountered
systems' and
developmental
and
organisms
the book, as we look at different
embarking on a study of
should be regardedas a conceptualtool kit, essentialfor
development.
Genescontroldevelopmentbycontrollingwhereandwhenproteinsaresyntheactivity setsup intracellusized,and many thousandsof genesare involved. Gene
genes'
and between proteins
and
proteins
lar networks of interactions between
one ofthese ProPerand proteins, that confer on cells their particular Propefties.
in specifiedways'
cells
other
to
ties is the ability to communicatewith and respond
it is t hes ece | | -c e l l i n te ra c ti o n s th a td e t ermi nehow theembryodevel ops;no
developmentalprocesscanthereforebeattributedtothefunctionofasinglegene
information on developor single protein. The amount of genetic and molecular
mentalProcessesisnowenolmous.Inthisbook,wewillbehighlyselectiveand
into the mechanismsof
describeonly those molecular details that give an insight
developmentand illustrate general principles'
of pattern'
t.E Developmentinvolvescell division,the emergence
changein form, cell differentiation, and growth
structules from an initiauy
Developmentis essentiallythe emergenceof organized
r,erysimplegroupofcells.Itisconvenienttodistinguishfivemaindevelopmental
influence one another
pr*.rr"r, even though in reality they overlap with and
considerablY.
II
13
14
1 : HIS T ORY A N D B ASIC CONCEPTS
Fig.1.14 LightmicrographofXenopus
eggs
afterfour celldivisions.Scalebar= 1 mm.
Photogroph
courtesy
ofl. Slock.
The first is the processknown as cleavage,which is the division of the fertilized
egg into a number of smaller cells (Fig. 1.14).Unlike the cell divisions that take
place during cell proliferation and growth of a tissue, there is no increasein cell
massbetween eachcleavagedivision. The cell rycles during cleavageconsistsimply
of phases of DNA replication, mitosis, and cell division, with no intervening
stageof cell growth. We shall discussthe various types of cell rycle in Chapter 12.
The cleavage stage of embryogenesis thus rapidly divides the embryo into a
number of cells, each containing a copy of the genome.
Patternformation is the processbywhich a spatial and temporal pattern of cellular activities is organized within the embryo so that a well ordered structure develops. In the developing arm, for example, pattern formation is the processthat
enablescells to 'know' whether to make an upper arm or fingers, and where the
muscles should form. There is no single universal strategy or mechanism of
patterning; rather, it is achieved by a variety of cellular and molecular rnechanisms
in different organismsand at different stagesof development.
Pattern formation initially involves laying down the overall body plan-defining
the main body axes of the embryo so that the head (anterior) and tail (posterior)
ends, and the back (dorsal) and underside (ventral) are specified. Most of the
animals we shall deal with in this book have a head at one end and a tail at the
other, with the left and right sidesof the body being bilaterally synrmetrical-that
is, a mirror image of each other. In such animals, the main body axis is the anteroposterior axis, which runs from the head to the tail. Bilaterally s5rmmetrical
animals also have a dorso-ventral axis, running fiom the back to the belly.
A striking feature of these axes is that they are almost always at right angles to one
another and can thus be thought ofas making up a systemofcoordinates on which
any position in the body could be specified(Fig.1.15).In plants, the main body axis
runs from the growing tip (the apex)to the roots and is known as the apical-basal
axis. Plants also have radial symmetry with a radial axis running from the center
of the stem outwards. Even before the body axesbecome clear, eggsand embryos
often show a distinct polarity, which in this context means that one end is distinguishable from the other in its structure or properties. Many individual cells in
the developing embryo also have polarity, with one end of the cell structurally or
functionally different from the other.
The next stagein pattern formation in animal embryos is allocation of cells to
the different germ layers-the ectoderm,mesoderm,and endoderm (Box 18, p. 15).
During further pattern formation, cells of these germ layers acquire different
0@
lM
iflw
nMl
dr
I
j$@
m
Itr
geembryo
,Y.laevirtailbud.sta
@
il!M'
(back)
Donal
rffi0
,u
itE
th
,:
Anterior
(head)
Posterior
(tail)
rym
Fig.1.15 Themainaxesof a developing
embryo.Theantero-posterior
axisandthe
dorso-ventral
axisareat rightanglesto one
another,
asin a coordinate
svstem.
m
g&
,tN
(front)
Ventral
im@
ilh
A C ON C EPT U AL T O O L K I T
Box 1B
Germ laYers
The concept of germ layers is useful to
distinguish between regions of the early
embryo that give rise to quite distinct types
of tissues.lt appliesto both vertebratesand
All the animalsconsideredin
invertebrates.
.c
ri
-':
this book, except for the coelenterateHydro'
are triploblasts,with three germ layers:the
endoderm,which givesriseto the gut and its
derivatives,suchasthe liverand lungsin vertebrates:the mesoderm,whichgivesriseto the
skeleto-muscularsystem, connectivetissues'
and other internalorganssuch as the kidney
and heart; and the ectoderm,which givesrise
to the epidermisand nervoussystem'These
The boundearlyin development'
arespecified
befuzzy
can
layers
different
ariesbetweenthe
neural
The
exceptions'
and there are notable
crestin vertebrates,for example,is ectodermal
in origin but gives rise both to neuraltissue
and to some skeletalelements,which would
in origin'
mesodermal
normallybe considered
-jentitiessothatorganizedspatialpattelnsofcelldifferentiationemerge,suchas
and the
-re arrangement of skin, muscle, and cartilage in developing limbs'
In the earliest stagesof pattern
_-rangementof neurons in the nervous system.
::rmation,differencesbetweencellsarenoteasilydetectedandprobablyconsist
changein activity of a very few genes'
-: subtle diffbrencescausedby a
process is change in form' or
The third important developmental
changesin three-dimenrorphogenesis (Chapter7). Embryosundergo remarkable
.lonalform_youneedonlylookatyoufhandsandfeet.Atcertainstagesindevelchanges in form' of which
, lment, there are characteristic and dramatic
undergo gastrulation'
embryos
all animal
; astrulatlon is the most striking. Almost
During gastrulaplan
emerges.
::ring which the gut is formed and the main body
such as the
animals
in
move inwards and'
-rn, cells on the outside of the embryo
into
blastula
hollow spherical
,.a urchin, gastrulation even transforms a
in
Morphogenesis
gut (Fig. 1.16).
. gastrulawith a hole through the middle-the
of
cells
the
of
Most
embryos can also involve extensive cell migration'
tissue
-rrmal
a
from
-:e human face, fbr example, are derived from cells that migrated
on the back of the embryo'
--Lledthe neural crest' which originates
Thefourthdevelopmentalprocesswemustconsiderhereisecl|differentiation,in
.hjchcellsbecome,.*.*,"uyandfunctionallydifferentfromeachother,ending
is a
as blood' muscle' or skin cells' Differentiation
-; as distinct cell types such
=adualplocess,cellsoftengoingthroughseveraldivisionsbetweenthetimeat
differentiated (when
.'nich they start differentiating and the time they are fully
,-necelltypesstopdividingaltogether)'andisdiscussedinChapter8'Inhumans'
least 250 clearly distinguishabletypes ofcell.
---:efertilized egg gives rise io
"t
r
lnsects
l5
16
.
1: HlsroRYANDBAslccoNcEPTs
Fig. 1.16 Gastrulationin the seaurchin'
GJrulation transformsthe sphericalblastula
into a structurewith a holethroughthe middle'
has
the gut. Theleft-handsideof the embryo
beenremoved.
w#M#
Patternformationandcelldifferentiationarevelycloselyinterrelated,aswe
canseebyconsideringthedifferencebetweenhumanarmsandlegs.Bothcontain
tle
cartilage' bone' skin' and so on-yet
exactly the same tyPes of cel-muscle'
different' It is essentially pattem
pattern in which they are arranged is clearly
elephants and chimpanzees'
formation that makes us different from
growth
in size' In general there is little
The fifth Processis growth-the increase
of tlre
form
and
pattern
and the basic
during early embryonic development
embryoislaiddownonasmallscale,alwayslessthanafewmillimetersinextenl
in a variety of ways: cell multiplication'
Subsequent$owth can be brought about
increaseincellsize,anddepositionofextracellularmaterialssuchasboneand
in that differences in growth rates
shell. Growth can also m Lorphogenetic
body' can generate changes in the overall
between organs, or between parts ofthe
see.inmore detail in Chapter 12'
shapeof the embryo (rig. r'fi), as we shall
neither independent of each other nor
These five a"rrutop*"L"1 p'ote's"s are
however' one can think of pattern formastrictly sequential. In very geileral terms'
#
tioninearlyd"rrulopm"ntspeci$'ingdifferencesbetweencellsthatleadtochanges
But in any real developing system there
in form, cell differenti"tor' *i, gro-ttt
of events'
will be many twists and turns in this sequence
gene action
1.e Cell behavior providesthe link between
and develoPmental Processes
tothe.svnth*it:fp1,"1:-*::Ti":ti|"t;1*:T
leads
expressionwithincells
Gene
J;;;;;"'""Jo"i"""r'whichinturnl"t"'**,tl:'::l::it^"::T:'T:
ffi "fi ffi;il
ffi;ffi
nt.t:l"."*-:tv*::*,:?:::':l;
pattern
;; ;;; *o""t
giventime, which is reflectedin its molecular
ft:;il;
liifili"i""-;;;;i*ffi
any
";roteins*"1'":"ll'-1'-Y-".^:Tt::"-?::3
cellsandtheirprogenyundelgomanychangesinstateasdevelopmentProglesses.
othercategoriesofcellbehaviorthatwillconcemusareintercellularcommunicachangesin cell shape and cell movement'
tion, also known as cell_cellsignaling,
cell proliferation, and cell death'
early development are essential for
Changing Patterns of gene activity during
that determine their future behavior
pattern formation. fn"y-gi* cells iientities
And' as we saw in the example of
and lead evenflrally to thJir nnat differentiation'
Fig. f .17 The human embryo changesshape
gro*r. Fromthe time the bodyplaniswell
"Jia
at 8 weeksuntil birth the embryo
established
panel)'
in
increases lengthsometenfold(upper
to the
head
whilethe relativeproPortionofthe
Asa
panel)'
(lower
restof the bodydecreases
Scale
changes'
embryo
ofthe
result,the shape
1983'
K'L':
Moore,
=
After
cm.
10
bar
inductionbythesp"**,,'o.ganizer,thecapacityofcellstoinfluenceeachother's
By their
.",poiaing to signals is crucial for development'
fate by producing
"rrai" ti"p"' cJh generate the physical forces that bring about
movement or change
a sheet of cells into a tube' as happens
morphogenesis Fig. f .rS)' ftt" t"-"*te
-of
irl,Xutoltusandothervertebratesduringformationoftheneuraltube(seeBoxlA,
generatedby cells changing their shape
pp. 4-5), is the result ofcontractile forces
atcertainpositionswithinthecelllayer.Animportantfeatureofcellsurfacesis
thepresenceofadhesiveproteinsknownascell-adhesionmoleculesthatserve
K IT
A C ON C E P T U ATOOL
L
!we
rhin
t tle
ftern
rwth
I the
tPnt.
tion,
and
ales
Erall
'nor
nTte-
ryes
here
dar
mic
:Il a
dar
mic
tes.
i:aEBt,
. for
rior
eo f
cr's
kir
trrt
ETIS
IA
Tle
nis
r5e
rr
a variety of functions: they hold cells together in tissues,they enable cells to sense
the nature of the surrounding extracellular rnatrix, and they sewe to guide
mrgratory cells such as the neural crest cells ofvertebrates, which leavethe neural
tube to form structures elsewhere in the body'
Later in development, gfowth involves cell proliferation, which can also influence final form, as parts of the body glow at different rates. The death of cells,
known as programmed cell death or apoptosis, is also an intrinsic part of the
developmental process; cell death in developing hands and feet helps to form
the fingers and toes from a continuous sheet of tissues.We can therefore describe
and explain developmental processesin terms of how individual cells and
groups of cells behave. Becausethe final structures generated by development are
themselvescomposedof cells, explanations and descriptions at the cellular level
can provide an account of how these adult structures are formed.
Since development can be understood at the cellular level, we can pose the
question of how genes control development in a more precise form. We can now
ask how genesare controlling cell behavior. The many possible ways in which a cell
can behave therefore provide the link between gene activity and the morphology
of the adult animal-the final outcome of development. cell biology provides
the means by which the genotype becomes translated into the phenotype.
r.1o Genescontrol cell behavior by specifyingwhich proteins are made
What a cell can do is determined very largely by the proteins plesent within it.
The hemoglobin in red blood cells enables them to transPort oxygen; skeletal
muscle cells are able to contract because they contain contractile strucfiIres
composed of the proteins myosin, actin, and tropomyosin, and othef musclespecific proteins. All these are vely special, or 'luxury" proteins that are not
involved in the housekeeping activities that are common to all cells and keep them
alive and functioning. Housekeeping activities include the production of energy
and the metabolic pathways involved in the breakdown and synthesis of molecules
necessaryfor the life of the cell. Although there are qualitative and quantitative
variations in housekeepingproteins in different cells, they are not important players
in developrnent. In development we are concerned primarily with those luxury or
tissue-specificproteins that make cells different from one another.
Genes control development mainly by speci8ring which proteins are made in
which cells and when. In this sensethey are passiveparticipants in development,
compared with the proteins for which they code, which are the agents that directly
determine cell behavior, including which genes are exPressed.To produce a particular protein its gene must be switched on and transcribed into messengerRNA
(mRNA);the mRNA must then be translated into protein. Both these processes
are under severallayers ofcontrol, and translation does not automatically follow
transcription. Fig. 1.19 shows the main stagesin gene expression at which the
production of a protein can be controlled. For example, nRNA may be degraded
before it can be exported from the nucleus. Even if it reachesthe cytoplasm, its
translation may be inhibited there. In the eggsof many animals, preformed nRNA
is prevented from being ffanslated until after fertilization. Another factor determining which proteins are produced is RNAprocessing.The initial RNAtranscripts
of many genes in eukaryotes can be spliced in different ways to give rise to two or
mofe diffefent mRNAs; thus a single gene may be able to produce a number of
different proteins with different properties.
Even if a gene has been transcribed and the mRNA translated, the protein
may still not be able to function. Many newly synthesized proteins require further
17
*
Fig. 1.18 Localizedcontractionof particular
cellscan causea whole sheetof cellsto fold.
of a lineof cellsat theirapicesdue
Contraction
elements
to the contractionof cytoskeletal
a furrowto formin a sheetof epidermis.
causes
18
.
l: HtsroRyANDBAstccoNcEprs
Fig'r'19 Geneexpressionandproteinsynthesis.Aprotein-codinggenecomprisesastretchof
DNAthatcontains
a coding
region,
whichcontains
theinstructions
formaking
theprotein,
and
adjacent
controlregions-promoter
andenhancer
regions-atwhichthegeneisswitched
onoroff.
Thepromoter
region
isthesiteatwhichRNApolymerase
bindsandstartstranscribing.
Theenhancer
regions
maybethousands
of basepairsdistantfromthepromoter.
Transcription
ofthegeneintoRNA
(1) maybeeither
stimulated
orinhibited
bytranscription
factors
thatbindto promoter
andenhancer
regions'
TheRNAformedbytranscription
isspliced
to remove
introns(yellow)
andprocessed
within
thenucleus
(2)to produce
mRNA
thatisexported
to thecytoplasm
(3)fortranslation
intoproteinat
theribosomes
(4).controlof geneexpression
andproteinsynthesis
occurs
mainlyatthelevelof
transcription
butcanalsooccurat laterstages.
Forexample,
mRNA
maybedegraded
beforeit canbe
translated'
If it isnottranslated
immediately
it maybestoredin inactive
forminthecytoplasm
for
translation
atsomelaterstage.
Someproteins
require
post-translational
modification
(5)to become
biologically
active.
post-translationalmodification before they acquire
biological activity. Reversible
post'translational modifications such as phosphorylation
can also significantly
alter protein function. Alternative RNA splicing and post-translational
modification
together mean that the number of functionally different proteins
that can be
produced is considerablygreater than the number
ofprotein-coding geneswould
indicate-perhaps as much as 10 times more.
some genes,such as those for ribosomal RNAs(rRNAs)and
transfer RNAs(tRNAs),
do not code for proteins; in this casethe RNAsthemserves
are the end products.
A recently discoveredcategoryof genesis that for microRNAs(miRNAs),
small RNA
moleculesthat inhibit the translation of specificmRNAs(seechapter
5, Box 58, p.7971.
some microRNAs are known to be invorved in gene regulation
in development.
An intriguing question is how many genes out of the total genome
are deveropmental genes-that is, genes specifically required for
embryonic deveropment.
This is not an easy estimate to make. In a few particularly
well studied caseswe
have rough estimates of the minimum number of genes
involved in a particular
aspectof development.In the early developmentof Drosophilaat
least 60 genesare
directly involved in pattern formation up to the time that
the embryo becomes
divided into segments.rn caenorhabditis,
at least 50 genesare neededto speci$ra
small reproductive structure known as the vulva. All these
are quite small numbers
comparedto the thousandsof genesthat are active at the
,.*" ti-"; someof these
are essentialto development in that they are necessary
for maintaining life, but
provide no or little information that influences the
course of development. Recent
studies have shown that many genes change their activity
during deveropment,
and as the total number of genesin the nematode and Drosophilais
around 19,000
and 20'000, respecfively,the total number ofgenes involved
in development is many
thousands.A systematicanalysis of nearly 9o%of the nematod"',
g"r", showed
that 7722 genes,at least, were involved in deveiopment.
Developmentar genes typically code for proteins involved
in the regulation of
cell behavior-receptors, growth factors,intraceilular signalingproteins,
and generegulatory proteins. Many ofthese genes,especiallythose
for receptorsand signaling molecules,are used throughout an organism's life,
but others are active only
during embryonic development.
r.tt The expressionof developmentalgenesis underthe
control
of complex control regions
All the somatic cells in an embryo are derived from the fertilized
egg by successive
rounds of mitotic cell division. Thus, with rare exceptions, they
all contain identical
TooL Krr
A coNcEPTUAL
r
dtrf
'RNA
lrc€f
!t'n
nat
nbe
r
IIE
ible
ntly
tion
nbe
urld
IAs),
ECtS.
TNA
1e7l,.
.
lopENL
I \A/e
dar
I are
mes
ify a
bers
ke
but
rFnt
Ent,
Ino
]tny
red
nof
!ne.
palmly
,Eeneticinformation, the same as that in the zygote. The differences between cells
must therefore be generatedby differences in gene activity that lead to the synthesis
ofdifferent proteins. Turning the correct geneson or offin the colTectcells at the
€orrect time becomesthe central issue in development.The genesdo not provide
a blueprint for development but a set of instructions. Key elements in regulating
tie readout of these instructions are the control regions located adjacent to most
tuxury genes and developmental genes.These are acted on by gene-regulatory
proteins, or transcription factors, which switch genes on or off by, respectively,
activating or repressing transcription. Some gene-regulatory proteins act by
hinding directly to the control regions (seeFig. 1.19)whereas others interact with
tmnscription factors already bound to the DNA.
Developmental genes are highly regulated to ensure they are switched on only
at the right time and place in development. This is a fundamental feature of development. To achieve this, they usually have extensive and complex control regions
composedof one or more regulatory modules, known as cis-regulatorymodules
tthe 'cis'refers to the fact that the regulatory module is on the sameDNA molecule
as the gene it controls). Each module contains multiple binding sites for different
ranscription factors, and the combination of factors that binds determines
whether the gene is on or off. On average,a module will have binding sites for four
to eight different transcription factors.
Different genes can have the same control module, which usually means that
tbey will be expressedtogether, or different genesmay have modules that contain
some, but not all, of the binding sites in cortmon, introducing subtle differences
in the timing or location of expression.A gene with more than one regulatory
module can be expressedat different times and placesduring development, depending on which module predominatesat any given time in directing gene expression.
Thus, an organism's genes are linked in complex interdependent networks
of expressionthrough their regulatory modules and the proteins that bind to them.
Common examples of such regulation are positive feedback and negative
feedbackloops in which a transcription factor respectivelypromotes or represses
the expressionof a genewhose product maintains that gene expression(Fig.1.20).
Lrdeed,the network ofinteractions between gene-regulatorymodules that has been
describedfor early sea urchin development is quite bewildering at first, or even
second,sight, as we shall seein Chapter 5.
As all the key steps in development reflect changes in gene activity, one might
be tempted to think of development simply in terms of mechanismsfor controlling gene expression.But this would be highly misleading. For gene expressionis
only the first step in a cascadeof cellular Processesthat lead via protein synthesis
to changesin cell behavior and so direct the course of embryonic development.
To think only in terms of genesis to ignore crucial asPectsof cell biology, such as
change in cell shape, that may be initiated at several steps removed from gene
activity. In fact, there are very few caseswhere the complete sequenceof eventsfrom
gene expression to altered cell behavior has been worked out. The route leading
from gene activity to a structure such as the five-fingered hand may be tortuous.
t.rz Developmentis progressiveand the fate of cellsbecomesdetermined
at different times
dlve
tl:al
II
As embryonic development proceeds,the organizational complexity of the embryo
becomes vastly increased over that of the fertilized egg. Many different cell types
are formed, spatial patterns emerge, and there are major changesin shape.All
this occurs more or less gradually, depending on the particular organism. But, in
19
Postitivefeedback
adivator
I
lr'J r-'
w.@.
gene2
gene
l
adivator
I
lrlff"
WjW
gene4
gene3
W
tr
Fig.1.2O Simplegeneticfeedbackloops.
Top:gene'l isturnedon byanactivating
factor(green);
the proteinit
transcription
gene2. TheProtein
produces
(red)activates
productof gene2 (blue)not onlyactson
gene1,
furthertargets,but alsoactivates
feedback
loopthatwill keep
forminga positive
genes1 and2 switched
on evenin the absence
Bottom:whenthe
of the originalactivator.
productof the geneat the endof the pathway
feedback
inhibitsthe firstgene,a negative
loopisformed.Arrowsindicateactivation;
inhibition.
barredlineindicates
20
.
r: HrsroRYANDBAstccoNcEPTs
Fig. 1.21 Thedistinction betweencell fate,
determination,and specification.In this
idealizedsystem,regionsA andBdifferentiate
into two differentsortsof cells,depictedas
Thefatemap(firstpanel)
andsquares.
hexagons
lf cells
showshowtheywouldnormallydevelop.
from regionBaregraftedinto regionA andnow
developasA-typecells,the fate of regionBhas
(secondpanel).By
notyet beendetermined
determined
if regionBcellsarealready
contrast,
whentheyaregraftedto regionA, theywill
developasBcells(thirdpanel).Evenif Bcellsare
in that
theymaybespecified,
not determined,
theywillform Bcellswhenculturedin isolation
fromthe restof the embryo(fourthpanel).
general, the embryo is first divided up into a few broad regions, such as the future
germ layers (mesoderm,ectoderm, and endoderm).Subsequently,the cells within
these regions have their fates more and more finely determined' Mesoderm,
for example, becomesdifferentiated into muscle cells, cartilage cells, bone cells,
the fibroblasts of connective tissue, and the cells of the dermis of the skin.
Determination implies a stable changein the internal state of a cell, and an alteration in the pattern of gene activity is assumed to be the initial step, leading to a
change in the proteins produced in the cell.
It is important to understand clearly the distinction between the normal fate
of a cell at any parLicular stage,and its state of determination. The fate of a group of
cells merely describes what they will normally develop into. By marking cells
of the early embryo one can find out, for example, which ectodermal cells will
normally give rise to the nervous system, and of those, which to the retina in
parLicular. However, that in no way implies that those cells can only develop into
a retina, or are already committed, or determined, to do so.
A group ofcells is called specified if, when isolated and cultured in the neutral
environment of a simple culture medium away from the embryo, they develop
more or less according to their normal fate (Fig. 1.21).For example, cells at the
animal pole of the amphibian blastula (seeBox 1'{, pp. 4-5) are specifiedto form
ectoderm, and will form epidermis when isolated. Cells that are specified in this
technical senseneed not yet be determined, for influences from other cells can
change their normal fate; if tissue from the animal pole is put in contact with cells
from the vegetalpole, the ani.malpole tissuewill form mesoderminsteadof epidermis. At a later stage of development, however, the cells in tJre animal region have
become determined as ectoderm and their fate cannot then be altered. Tests for
specification rely on culturing the tissue in a neutral environment lacking any
inducing signals,and this is often difficult to achieve.
The state of determination of cells at any developmental stage can be demonstrated by transplantation experiments. At the blastula stage of the amphibian
K tT
L
A C ON C E P TUATOOL
prcsumPwe
eyeregron
presqmptivs
eprderm6
0resumDllve
'endoilerm
presumplve
plate
neural
tf-
I
Iailbud-stageembrYo
presumptive
\mesoderm
blaslopore
fitg.1.22 Determinationof the eye region
with time in amphibiandevelopment.lf the
thatwill normallygiverise
regionof the gastrula
to aneyeisgraftedintothetrunkregionof a
neurula(middlepanel),the graftforms
suchas
typicalof its newlocation,
structures
notochordandsomites.lf, however'the eye
regionfrom a neurulaisgraftedintothe same
asaneyeJike
site(bottompanel),it develops
it has
stage
later
at
this
since
structure,
becomedetermined.
bre
rhin
ErIn
plls,
*infterto a
frte
|Pof
rells
rwill
E io
rinto
ffiaI
rlop
t the
brm
r this
D CAII
rcells
edderrhalr
ts for
g eny
lEloIF
frian
ri
21
side of
embryo, one can graft the ectodermal cells that give rise to the eye into the
is,
that
new
their
to
Position;
de body and show that the cells develop according
At
this
(Fig.
7.22l|.
somites
into rnesodermalcells like those of the notochord and
early stage,their potential for development is much greater than their normal fate'
Itrowever,if the same operation is done at a later stage,then the future eye region
yill form structules tyPical of an eye. At the earlier stage the cells were not yet
determined as eye cells, whereaslater they had become so'
It is a general feature of development that cells in the early embryo are less
more and
uarrowly determined than those at later stages;with time, cells become
that
determination
assume
We
potential.
developmental
more restricted in their
fovolves a change in the genes that are expressedby the cell and that this change
rmxes
or restricts the cell's fate, thus reducing its developmentaloptions.
We have already seen how, even at the two-cell stage,the cells of the seaurchin
rmbryo do not seem to be determined. Each has the potential to genefate a whole
much
uE\vlarva (seeSection1.3).Embryoslike these,where the potential of cells is
Vertebrate
regulative'
termed
are
fate,
&Featerthan that indicated by their normal
cmbryos are among those capable of considerable regulation. In contrast, those
to
rmbryos where, from a very early stage, the cells can develop only according
has
mosaic
the
term
1.3,
rfreir early fate are termed mosaic,As we saw in Section
rLbng history. It describes eggs and embryos that develop as if their pattern of
ffiuturedevelopment is laid down very early, even in the egg, as a mosaic of differrnt molecules. The different parts of the embryo then develop quite independently
demardeach other. In such embryos, cell interactions may be quite limited. Ttre
oafion between the regulative and mosaic strategies of development is not always
frarp, and in part reflects the time when determination occurs; it occurs much
emlier in mosaic systems.
22
.
r: HrsroRYANDBAstccoNcEPTs
The difference between regulative and mosaic embryos reflects the relative
importance of cell-cell interactions in each system. The occurrence of regulation
absolutely requires interactions between cells, for how else could normal develop
ment take place and deficiencies be recognized and restored? A truly mosaic
embryo, however, would in principle not require such interactions. No purely
mosaic embryos are known to exist.
1.r3 Inductiveinteractionscan make cellsdifferent from eachother
Fi9.1.23 An inducingsignalcan be
transmitted from one cell to anotherin
three mainways.Thesignalcanbea diffusible
witha receptor
on the
molecule.
whichinteracts
(secondpanel),or the signal
targetcellsurface
canbe producedby directcontactbetween
proteinsat the cell
two complementary
(thirdpanel).lfthe signalinvolves
surfaces
it maypassdirectlyfrom cell
smallmolecules
to cellthroughgapjunctionsin the plasma
(fourthpanel).
membrane
Making cells different from one another is cenffal to development. There are
numerous examples in development where a signal from one group of cells influencesthe development of an adjacent group of cells. This is known as induction
and the classic example is the action of the Spemann organizer in arnphibiens
(see Section 1.4). Inducing signals may be propagated over several or even many
cells, or be highly localized. The inducing signals from the amphibian organizer
affect many cells, whereas other inducing signals may passfrom a single cell to its
immediate neighbor. T\uo different types of inductions should be distinguished:
permissiveand instructive. Permissiveinductions occur when a cell makes only
one kind of response to a signal, and makes it when a given level of signal is
reached. By contrast, in an instructive induction, the cells respond differently to
different concenffations of the signal.
There are three main ways in which inducing signals may be passedbetween
cells (Fig. 1.23).First, the signal can be transmitted through the extracellular space,
usuallybymeans of a secreteddiffirsible molecule. Second,cells mayinteract directly
with each other by means of molecules located on their surfaces. In both these
cases,the signal is generally receivedby receptor proteins in the cell membrane
and is subsequently relayed through intracellular signaling systems to produce
the cellular response.Third, the signal may passfrom cell to cell directly. In animal
cells, this is through gap junctions, which iue specializedprotein pores in tJre
apposedplasma membranes providing direct channels of communication between
the rytoplasms of adjacent cells through which small molecules can pass. Plant
cells are connected by strands of cytoplasm called plasmodesmata,through which
even quite large molecules such as proteins can passdirectly from cell to cell.
In the caseof signaling by a diffirsible molecule or by direct contact, the signal
is received at the cell membrane. If it is to alter gene expression in the nucleus, the
signal has to be transmitted from the membrane to the cell's interior. This process
is known generally as signal transduction, and is carried out by relays of
intracellular signaling molecules that are activated when the extracellular signal
molecule binds to its receptor. Intracellular signaling proteins and small-molecule
second messengerssuch as cyclic AMP interact with one another to transmit ti.e
signal onward in the cell. Activation and deactivation of the pathway components
by phosphorylation is an important feature of most signaling pathways. In the case
of development, most of the signaling we are concerned with leads to gene activation or repression, but signaling pathways are also used to temporarily alter enzyme
activity and metabolic activity within cells and to relay nerve impulses. The variety
of signals a cell may be receiving at any one time are integmted by cross-talk
between the different signaling pathways to produce an appropriate response.
An important feature of induction is whether or not the responding cell is
competent to respond to the inducing sigrral.This competencemay depend, for
example,on the presenceof the appropriate receptor and transducing mechanism,
or on the presenceofthe particular transcription factors neededfor gene activatiorl
A cell's competence for a particular response can change with time; for example,
K IT
A C ON C E P TUATOOI
L
htive
trion
rclop
psaic
urely
t are
influrtion,
bians
many
rnizer
I to its
bhed:
s only
pal is
ltly to
tween
sPace,
irectly
, these
brane
mduce
rnimal
in the
ltween
- Plant
which
t: signal
ns, the
lrOCeSS
h:n of
: signal
olecule
nit the
Etrents
he case
:activanzyme
Yadety
rs-talk
te.
; cell is
nd, for
nnism,
iration.
emple,
IT
rhe Spemann organizer can induce changes in the cells it affects only during
a restricted time window
[n embryos, it seemsthat small is generally beautiful where signaling and Pattem
formation are concefned. Whenever a Pattern is being specified, the size of the
group of cells involved is barely, if ever, greatel than 0.5 mm in any direction: that
is, some 50 cell diameters. Many patterns ale specified on a much smaller scaleand
hvolve just tens or a few hundreds of cells. This means that the inducing signals
involved in pattem formation reach over distances of the order of only 10 times
a cell diameter. The final olganism may be very big, but this is almost entirely due
to growth of the basic Pattern.
t.t4 The responseto inductive signalsdependson the state of the cell
Inductive signals can alter how the induced cells develop. They can thus be
regarded as providing the cells with instnrctions as how to behave.After receiving
an inductive signal, a cell usually then develops autonomously that is' without
new signals from another cell, for some time' It is important to realize that the
Fesponseto inductive signals is entirely dependent on the current state ofthe cell.
llot only does it have to be competent to respond, but the number of possible
responsesis usually very limited. An inductive signal can only select one response
from a small number of possible cellular responses. Thus instructive inducing
signals would be more accurately called selective signals. A truly instructive signal
would be one that provided the cell with entirely new information and capabilities'
by providing it with, for example, new genes,which is not thought to occtll during
development.
The fact that, at any given time, an inductive signal selects one out of several
possible responseshas several imPortant implications for biological economy' On
ttre one hand, it means that different signals can activate a particular gene at different stages of development: the same gene is often turned on and off repeatedly
during development. On the other, the same signal can be used to elicit different
responsesin different cells. A particular signaling molecule' for example, can act
on severaltypes ofcell, evoking a characteristicand different responsefrom each,
depending on their developmental history. This enables the fixed set of genes and
signals to be related to each other in a variety of different combinations' effectively
expanding the number of developmentally distinct signals and responsesthat can
be made by cells. As we will see in future chapters, evolution has been lazy with
respect to this aspect of development, and the same basic intracellular signaling
pathways are used again and again for different purposes'
1.rs Patterningcan involve the interpretation of positional information
One general mode of pattem formation can be illustrated by considering the
patterning of a simple non-biological model-the French flag (Fig. 7.24).T\e French
flag has a simple pattem: one-third blue, one-third white, and one-third red, along
just one axis. Moreover, the flag comes in many sizesbut always the same Pattern'
and thus can be thought of as mimicking the capacity of an embryo to regulate.
Given a line of cells, any one of which can be blue, white, or red, and given also
that the line of cells can be of variable length, what sort of mechanism is required
for the line to develop the pattern ofa French flag?
One solution is for the group of cells to acquire positional information' That is'
each cell acquires an identity, or positional value, that is related to its position
along the line with respect to the boundaries at either end' After they have
Fig.1.24 TheFrenchflag.
23
24
I: HIS T O RY AND BASIC CONCEPTS
acquired their positional values, the cells interpret this information by differentiating according to their genetic program. Those in the left-hand third of the line
will become blue, those in the middle third white, and so on.
Pattern formation using positional information thus implies at least two distinct
stages:first the positional value has to be specified with respect to some boundary,
and then it has to be interpreted. The separation of these two Processeshas an
important implication: it means that there does not need to be a set relation
between the positional values and how they are interpreted. In different circumstances,the same set of positional values could be used to generate the Italian flag
or another pattern. How positional values will be interpreted will depend on the
particular genetic instructions active in the group of cells and will be influenced by
their developmental history.
Cells could have their position specified by various mechanisms. The simplest is
based on a gradient of some substance. If the concentration of some chemical
decreasesfrom one end of a line of cells to the other, then the concentration of
that chemical in any cell along the line effectively specifies the position of the cell
with respect to the boundary (Fig. 7.25). A chemical whose concentration varies
and which is involved in pattern formation is called a morPhogen.In the caseof
the French flag, we assume a source of morphogen at one end and a sink at the
other, and that the concentrations ofmorphogen at both ends are kept constant
but are different from each other. Then, as the morphogen diftrses down the line,
its concentration at any point effectively provides positional information. If the
cells can respond to threshold concentrations of the morphogen-for example,
above a particular concentration the cells develop as blue, while below this concentration they become white, and at yet another, lower, concentration they become
red-the line of cells will develop as a French flag (seeFig. 1.25).Thresholds can
represent the amount of morphogen that must bind to receptors to activate an
intracellular signaling system, or concentrations of transcription factors required
to activate particular genes. The use of threshold concentrations of transcription
factors to speciSrposition is most beautifully illustrated in the early development
of Drosophila,as we shall see in Chapter 2. Other ways of speciffing positional
information are by direct intercellular interactions and by timing mechanisms,
as discussedin Chapter 4 and elsewherein the book.
The French flag model illusffates two important features of development in
the real world. The first is that, even if the lengttr of the line varies, the system
regulates and the pattern will still form correctly, given that the boundaries ofthe
system are properly defined by keeping the different concentrations ofmorphogen
constant at either end. The second is that the system could also regenerate the
complete original pattern if it were cut in half, provided tfiat the boundary concentrations were re-established. It is therefore truly regglative. We have discussedit
Fig. 1.25 The Frenchflag model of pattern formation. Eachcellin a lineof cellshasthe potential
gradientof some
to a concentration
to developasblue,white,or red.Thelineof cellsisexposed
positional
concentration
at that point.Each
value
defined
by
the
acquires
a
substance
andeachcell
intoblue,white,or red,
anddifferentiates
valueit hasacquired
celltheninterprets
the positional
that
flagpattern.Substances
geneticprogram,
thusformingthe French
according
to a predetermined
of
Thebasicrequirements
of cellsin thiswayareknownasmorphogens.
candirectthe development
at eitherendof the gradientmustremain
of substance
sucha systemarethat the concentration
to the system.Eachcellmustalso
differentfrom eachotherbut constant,thusfixingboundaries
containthe necessary
informationto interpretthe positionalvalues.Interpretationofthe positional
of morphogen.
to differentconcentrations
valueis basedupondifferentthresholdresponses
K IT
A C ON C E P TU ATOOL
L
trF
lne
TKT
ay"
ia n
tbn
IIIF
8ac
the
ilb"v
5t is
d{-el
no f
: cell
ries
ro f
t tie
lant
line,
f the
Eple,
tctnilrme
I €an
!e an
uircd
ption
ment
tioDal
isms,
nt in
Ftem
of tle
hogen
F the
flcen$ed it
E ttial
E
tacft
d.
Ef that
rnts of
il
bo
lbnal
IT
25
here as a one-dimensionalpatterning problem, but the model can easily be extended
to provide two-dimensionalpatterning (Fig.1.26).
br many casesit is still undear how positional information is specified-molphogen
diffirsion is but one of tJremechanisms tlrat have been discovered-but the evidence
for it is strong. However, we do not lrrow how fine-grained the differences in position are. For example, has every cell in the early embryo a distinct positional value
of its own and is it able to make use of this difference?
An important aspect of many patterning processesis the setting up of a
polarity in the plane of a sheet of cells. It is as if a direction arrow is Presentin all
the cells in the sheet making one end slightly different from the other, rather like
a compass needle pointing north. This type of polarity is known as planar cell
polarity and is discussed further in Chapters 2 and 7. One obvious example of
planar cell polarity is the hairs on the Drosopltilawing, which all point in the same
direction.
r.r6 Lateralinhibition can generatespacingPatterns
Many structures,such as the featherson a bird's skin, afe found to be more or less
regularly spaced with respect to one another. One mechanism that gives rise to
such spacing is called lateral inhibition (Fig. 1.27).Given a gloup of cells that all
have the potential to differentiate in a particular way, for example as feathers, it is
possible to regularly space the cells that will form feathers by a mechanism in
which cells that.begin to form feathers inhibit the adjacent cells from doing so.
This is reminiscent of the spacingof trees in a forest being causedby competition
for sunlight and nutrients. In embryos, lateral inhibition is often the result of the
differentiating cell secreting an inhibitory molecule that acts locally on the nearest cell neighbors to prevent them developing in a similar way.
Flg. 1.25 Positionalinformation can be used
to generatePatterns.A goodexample,as
shownhere,iswherethe peopleseatedin a
stadiumeachhavea positiondefinedbytheir
Eachpositionhasan
rowandseatnumbers.
aboutwhichcoloredcardto holdup,
instruction
andthismakesthe pattern.lf the instructions
werechanged,a differentpatternwouldbe
canbe
information
formed,andsopositional
usedto produceanenormousvarietyof patterns.
!.17 Localizationof cytoplasmicdeterminantsand asymmetriccell division
can make cells different from each other
Positional specification is just one way in which cells can be given a particular
identity. A separatemechanism is basedon cytoPlasmiclocalizationand asymmetric cell division (Fig.1.28).Asymmetric divisions are so called becausethey result in
daughter cells having properties different from each other, independently ofany
environmental influence. The properties of such cells therefore depend on their
lineageor line of descent,and not on environmental cues.Although some aspnmetric cell divisions are also unequal divisions in that they produce cells of different sizes, this is not usually the most important feature in animals; it is the
unequal distribution of cytoplasmic factors that makes the division asymmetric.
An alternative way of making the French flag pattern from the egg would be to
havechemical differences(representingblue, white, and red) distributed in the egg
fo the form of determinants that foreshadowed the French flag. When the egg
underwent cleavage, these cytoplasmic determinants would become distributed
:moilg the cells in a particular way and a French flag would develop. This would
require no interactions between the cells, which would have their fates determined
fum the beginning.
Although such extreme examplesof mosaic development are not known in nature,
there are well known caseswhere eggs or cells divide so that some cytoplasmic
determinant becomes unequally distributed between the two daughter cells and
frey develop differently. This happens at the first cleavage of the nematode egg,
ior exarnple, and defines the antero-Posterior axis of the embryo. The germ cells of
Fig.1.27 Lateralinhibition can give a spacing
pattern.Lateral
inhibitionoccurswhen
producean inhibitorthat
structures
developing
prevents
the formationof anysimilarstructures
to them,andthe structures
in the areaadjacent
becomeevenlyspacedasa result.
26
I: H IS T OR Y AND EASIC CONCEPTS
Drosoythilaare also specified by cytoplasmic determinants,
in this casecontained in
the cytoplasm located at the posterior end ofthe egg,
and in Xenoptsakey deter_
minant in the first stagesof embryonic development
is a protein t*o*
as vegT,
which is localized in the vegetal region of the fertilizea
ejg. In generar, however,
as developmentproceeds,daughter cellsbecomedifferent
from eachother because
of signals from other celrs or from their extracellular
environment, rather than
becauseof the unequal distribution of cytoplasmic determinants.
A very particular qpe of asymmetric division is that of stem
cells, which are cells
that are capable of repeated division and that produce
daughter cens, one of
which remains a stem ceil while the other differentiates
into a iariety of cell types
(Fig.1.29).stem cells that are capableofgiving rise
to all the cell types in the body
are present in embryos, while in adults, stem cells
of apparently more rimited
potential are responsiblefor the continual renewal
of tissuessuch asblood, epidermis,
and gut epithelium and for the replacement of tissues
such as muscle when required.
The difference in the daughter cells can be due either
to asyrnmetric distribution
of cytoplasmic determinants or to the effects of external
signals. The generation of
neurons in Drosophila,for exampre, depends on the
distribution of cytoprasmic
determinants in neuronal stem cells. Stem cells that can give
rise to all other types
of cells,but not to a complete embryo, are known as pluripotent.
The fertilized egg,
which can give rise to a complete embryo, is described
as totipotent.
t 'ra The embryo contains a generative rather than
a descriptive
Flg. 1.28 Cell division with asymmetric
of cytoplasmic determinants, lf a
partiotlar molecule is distributed unevenly in
distribution
the parent cell, cell dMsion will result in its being
shared unequally betrrveenthe cytoplasm ofthe
two daughter cells. The more localizedthe
cytoplasmic determinant is in the parental cell,
the more likelyit will be that one daughtercell
will receiveall of it and the other none,thus
producing a distinct difference between them.
program
All the information for embryonic development is
contained within the fertilized
egg' so how is this information interpreted to give rise
to an embryo?one possibility
is that the structure ofthe organism is somehow encoded
as a descriptive program
in the genome. Does the DNA contain a furl description
of the organism to which
it will give rise: is it a blueprint for the organism?
The answer is no. Instead, the
genome contains a program of instructions for making
the organism_a generative
program-in which the cytoplasmic constituents
of eggs and cells are essential
players along with the genes.
A descriptive program such as a blueprint or a plan
describes an object in some
detail, whereas a generative program describes how
to make an oulect. For the
same object the programs afe very different. consider
origami, the art of paper
folding. By folding a piece of paper in various directions
it is quite easy to make
a paper hat or a bird from a single sheet. To describe
in any detail the finar form of
the paper with the complex relationships between its parts
is really very difficult,
and not of rruch help in explaining how to achieve it.
Much rnore useful and easier
to forrnulate are instmctions on how to fold the paper.
The reason for this is that
simple instructions about folding have complex spatial
consequences.In deverop
ment' gene action similarly sets in motion a sequence
of events that can bring
about profound changesin the embryo. one can thus
think of the genetic information in the fertilized egg as equivalent to the
folding instructilns in origami;
both contain a generative program for making a particular
structure.
A further distinction can be made between the
embryo,s genetic program and
the developmental program of a particular cell
or group of ce[s. The genetic
program refers to the totality of information provided
by the genes, whereas a
developmental program may refer only to that part
of the genetic program that is
controlling a particular group of ce[s. As the embryo
develops, diFerent parts
acquire their own developmental programs as a result
of cell-cell interactions and
the activities of selectedsets of genes.Each celr
in the embryo thus has its own
developmental program, which may change as development
proceeds.
K IT
A C ON C E P TU ATOOL
L
t-r
p
ttE
gn"
dk
:d
p(5
odr
lrFd
Eis,
iedtbn
no f
IEic
t?es
.€'
ilized
bility
Eram
rhich
rL the
rative
rntial
lsome
irr the
'paper
r make
irm of
iltrcult,
I easier
i is tfiat
hreloP
n bring
nformarigami;
am and
genetic
hereas a
n that is
nt parts
ions and
l its own
il
t.tg The reliability of development is achievedby a variety of means
The development of embryos is remarkably consistent and reliable; you need only
look at the similar lengths of your legs, each of which developsquite independently
of the other for some 15 years (Chapter 9). Embryonic development needsto be reliable in the sensethat the adult organism can function properly. For example, both
wings of a bird must be very similar in size and shape if it is to be able to fly satisfactorily. How such reliability is achieved is an important problem in development'
Developrnent needs to be reliable in the face of fluctuations that occur either
within the embryo or in the external environment. Internal fluctuations include
small changesin the concentrations of molecules,and also changesdue, for example,
to mutations in genes not dfuectly linked to the development of the organ in
question. External factors that could perturb development include temPerature
and environmental chemicals.
One of the central problems of reliability relatesto pattern formation. How are cells
specifiedtobehaveinaparticularwayina partiorlarposition?TWowaysinwhichreliability is assured ale aPparent redundancy of mechanism and negative feedback'
Redundancyis when there are two or more ways of carrying out a particular process;
if one fails for any reason another will still function. It is like having two batteries in
your car instead of just one. True redundancy-such as having two identical genesin
a haploid genome with identical functions-is IaIe except for casessuch as rRNA'
where there can be hundreds of similar genes.Apparent redundancy, on the other
hand, where a given Processcan be specifiedby severaldifferent mechanisms,is probably one of the ways that embryos can achieve such precise and reliable results. It is
like being abte to draw a straight line with either a ruler oI a piece of taut string. This
type of situation is not true redundancy, but may give the impression of being so if
one mechanism is removed and the outcome is still apparently normal'
Negative feedback also has a role in ensuring consistency; here, the end product
ofa processinhibits an earlier stageand thus keepsthe level ofproduct constant
(see,for example, Fig. 1.20).The classicexample is in metabolism, where the end
product of a biochemical pathway inhibits one of the enzymes that act early in
the pathway. Yet another reliability mechanism relates to the complexity of the
networks ofgene activity that operate in development. There is evidence that these
networks are robust and relatively insensitive to small changes in, for example,
the rates of the individual processesinvolved.
1.20 The complexity of embryonic development is due to the complexity
of cells themselves
Cells are, in a way, more complex than the embryo itself. In other words, the
network of interactions between proteins and DNA within any individual cell
contains many more components and is very much more complex than the interactions between ttre cells of the developing embryo. However clever you think cells
are, they are almost always far cleverer. Each of the basic cell activities involved in
development,such as cell division, fesPonseto cell signals,and cell movements,
are the result of interactions within a population of many different intracellular
proteins whose composition varies over time and between different locations in the
cell. Cell division for example, is a complex cell-biological Program that takes Place
over a period of time, in a set order of stages,and requires the construction and
preciseorganization of specializedintracellular structures at mitosis.
Any given cell is expressing thousands of different genes at any given time.
Much of this gene expression may reflect an intrinsic pfoglam of activity'
27
Stemrells
a'@
-oo
b
"(_
oi 'o
io
Fig.1.29 Stemcells.Stemcells(S)arecells
andgiveriseto
that both renewthemselves
celltypes.Thus,a daughtercell
differentiated
arisingfromstemcelldivisionmayeither
developinto anotherstemcellor giveriseto a
quitedifferenttype of cell(X).lt mayalsodivide
to giveriseto two stemcellsor to two different
typesof cells(XandY).
28
.
r: HtsroRYANDBAslcc oNc EPTS
independent of external signals. It is this complexity that determines how cells
respond to the signals they receive; how a cell lesponds to a Particular signal
depends on its internal state. This state can reflect the cell's developmental
history-cells have good memories-and so different cells can respond to the same
signal in very different ways. We shall see many examples of the same signals being
used over and over again by different cells at different stages of embryonic
development, with different biological outcomes.
We have at present only a fragmentary picture of how all the genes and proteins
in a cell, let alone a developing embryo, interact with one another. But new
technologies are now making it possible to detect the simultaneous activity of
hundreds of genes in a given tissue. The discipline of systems biology is also beginning to develop techniques to reconstruct the highly complex signaling networks
used by cells. How to interpret this information, and make biological sense of the
patterns of gene activity revealed, is a massive task for the future.
!l il (4Ir
I l rfltlfllilm
riLI'nl|],tilft
1 l u , L rt g
;iilrlli:lglllllT
l1l-tudfl4tt
lll ""lltr'
ll ru,tTlN
-rl|lliilllm
t'(
tliill
--rmfln
q
iilul'l
1Lllflmli
iiiii|llrfiu0
ur{c,i 1
Surnrnary
iriNlllirlll
.tr
of cells.ThemajorProcesses
behavior
resultsfromthe coordinated
Development
iill|" m0
or
are cell division,Patternformation,morphogenesis
involvedin development
groMh.
Genes
cell migration,cell death,and
changein form, celldifferentiation,
andthus
aresynthesized,
controlcellbehaviorbycontrollingwhereandwhenProteins
processes.
cell biologyprovidesthe link betweengene action and developmental
in theirshape,
in the genestheyexpress,
Duringdevelopment,
cellsundergochanges
and in their
in the signalsthey produceand resPondto, in their rateof proliferation,
arecontrolledlargelyby the presof cellbehavior
Alltheseaspects
migratorybehavior.
enceof specificproteins;geneactivitycontrolswhich proteinsare made.Sincethe
somaticcellsin the embryogenerallycontainthe same geneticinformation,the
activityof selected
arecontrolledbythe differential
that occurin development
changes
to
setsof genesin differentgroupsof cells.Thegenes'controlregionsarefundamental
at
determined
becomes
fate
of
cells
progressive
the
process.
and
is
Development
this
is
embryo
usually
in
the
early
cells
of
potential
development
The
for
differenttimes.
much greaterthan their normalfate,but this potentialbecomesmore restrictedas
involvingsignalsfromonetissueor cell
proceeds.
Inductiveinteractions,
development
to another,areoneof the mainwaysof changingcellfateanddirectingdevelopment.
distributed
areunequally
comPonents
celldivisions,
in whichcytoplasmic
Asymmetric
means
of pattern
to daughtercells,can also makecellsdifferent.One widespread
positional
valuewith
cellsfirstacquirea
information;
generation
isthroughpositional
valuesby behavingin differrespectto boundaries
andthen interprettheirpositional
choosingoneor
thaninstructive,
signalsaremoreselective
ent ways.Developmental
other of the developmental
Pathwaysopen to the cell at that time. The embryo
for
program-it is morelikethe instructions
not a descriptive,
containsa generative,
includA varietyof mechanisms,
by paperfoldingthana blueprint.
makinga structure
areinvolvedin makingdevelopment
andnegativefeedback,
ingapparentredundancy
lieswithinthecells.
of development
remarkably
reliable.
Thecomplexity
All the informationfor embryonicdevelopmentis containedwithin the fertilized
egg-the diploidzygote.The genomeof the zygotecontainsa programof instrucprogram,
tionsfor makingthe organism.In the workingout of this developmental
of the egg and of the cellsit givesriseto areessential
the cytoplasmicconstituents
i
r" ll1ll
.i (
" lilLL
ifl lllll
. illllll
FU R T H ERRE A D IN G
EIIS
Fal
@1
tne
rioC
mic
Eins
ts\f,
tr of
EIF
Ets
trhe
I
3
F
yr
F
Ps
F
FF
playersalong with the genes.Strictly regulatedgene activity, by controllingwhich
proteinsare synthesized,and where and when, directs a sequenceof cellular
activitiesthat bringsabout the profoundchangesthat occurin the embryoduring
development.The major processesinvolvedin developmentare cell division,
cellmigration,celldeath,and
celldifferentiation,
patternformation,morphogenesis,
growth.All of theseprocesses
canbe influencedby communicationbetweenthe cells
of the embryo. Developmentis progressive,with the fate of cellsbecominglrnre
Cellsin the earlyembryousuallyhave
preciselyspecifiedasdwelopmentproceeds.
potential
for developmentthan is evident from their normal fate,
a much greater
and this enablesembryosto developnormallyevenif cellsare remwed, added,or
potentialbecomesmuch
transplantedto differentpositions.Cells'developmental
proceeds.
Manygenesareinvolvedin controllingthe
morerestrictedasdevelopment
complexinteractionsthat occurduringdevelopment,and reliabilityis achievedin
ofanimalsandplants,
ofgenescontrolthe development
a varietyofways.Thousands
and a full understandingof the processesinvolvedis far from beingachieved.While
basicprinciplesand certaindevelopmentalsystemsare quite well understood,there
arestillmanygaps.
FURTHER
READING
The origins of developmental biology
F
F"
Cole, F.l.: Eorly Theoriesof SexuolGmration. Oxford: Clarendon
Press. 1930.
Hamburger, Y.: TheHerttuge of Experimmtal Embryolagt: Hans
Spemannondthe OrganiTrr. NewYork: Oxford University Press'
fb
1988.
Milestones in D 6relof,ment
F-
ts
!lv
la
tr
FtFd
Fin
Fft
F
F{r
rp
lbt
l+
F*
I
t
I
F.
t*
Irr.
tIral
i
II
[http:/lriuww.nature.corn/milestones/development/index.htrn!
Needham, J.: A Histo'ry of Enfuryolog. Cambridge: Cambridge
University Press, 1959.
effects" in
and "assimilating
Sander, K.: "Mosaic worlf
after
embryogenesis: Wilhelm Roux's condusions
Arch. D4,. Biol. 1997,
disabling frog blastometes. Rotu<'s
2OO:237-239.
Sander, K.: Shaking a oonaept flens f)riesch and the varied
fates of sea uff:hin blastomeres. Rotn's Arch. Dett Biol. 7992'
z()li 265-267.
Wilson, E.B.: TheCel in Developmentond Heredity. New York:
Macmillan. 1896.
Wolpert, L.: Evolution of tlre cell theory. PhiI. Trors. Roy.Soc.
Lond.B 1995, 349t 227-233.
A conceptual tool kit
Alberts, 8., et aL: EssentialCeIIBiologt: An Introduchon to the
Moleaiar Biologr of the Ce'il.2nd edn. New York: Garland
Science. 2003.
splicing: increasing diversity
Graveley, B.R: Alt€mative
TrendsC'enet.2007, a7: 7oo-1o7.
in the proteornic wotlil
Howard, M.L., Davidson, E.H.: cis-Regulatory aontrol circtrits
in developrnent. Dst. Biol. 2oM,271: 109-118.
Jordan, J.D., Landau, E.M., Lyengar, R.: Signaling networks:
CeI] 2ooo, lO3:
the origins of ce[ular multitasking.
193-200.
and animal
rcgulation
Levine, M., Tjian, R: ltanscriptional
diversity. Nature 2OO3,424: 747 -157.
of corre mechanisms to generate
Nelson, WJ.: Adaptation
polarity. Notttre 2oo3, 422: 766-77 4.
Papin, JA., Hunter, T., Palsson, B.O., Subramanian, S.:
networks and
of cellular sip.alling
Reconstnrction
analysis of their properties. Nat Rev.MoI. Biol. 2oo5, 6l
99-777.
development?
wolpert, L.: Do we understand
266i 571-572.
Wolpert, L.: One hundred years of positional
TrendsGenet.1996. a2: 359-364.
Science1994'
information.
29