Gene Regulation of Flowering

Author(s): D. Bradley, R. Carpenter, R. Elliott, R. Simon, J. Romero, S. Hantke, S. Doyle, M.

Mooney, D. Luo, P. McSteen, L. Copsey, C. Robinson, E. Coen
Source: Philosophical Transactions: Biological Sciences, Vol. 339, No. 1288, Transgenic
Modification of Germline and Somatic Cells: Examples from Animals and Plants (Feb. 27,
1993), pp. 193-197
Published by: The Royal Society
Stable URL: [Link] .
Accessed: 09/07/2011 14:30
Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at .
[Link] JSTOR's Terms and Conditions of Use provides, in part, that unless
you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you
may use content in the JSTOR archive only for your personal, non-commercial use.
Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at .
[Link] .
Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed
page of such transmission.
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of
content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms
of scholarship. For more information about JSTOR, please contact support@[Link].

The Royal Society is collaborating with JSTOR to digitize, preserve and extend access to Philosophical
Transactions: Biological Sciences.

[Link]

Gene regulationof flowering

D. BRADLEY, R. CARPENTER, R. ELLIOTT, R. SIMON, J. ROMERO,
S. HANTKE, S. DOYLE, M. MOONEY, D. LUO, P. McSTEEN,
L. COPSEY, C. ROBINSON AN:D E. COEN
Department
of Genetics,AFRC Instituteof Plant ScienceResearch,JohnInnes Centre,ColneyLane,
NorwichNR4 7UH, U.K.

SUMMARY
A major change in the development of plants occurs upon floral induction. Meristems in certain
positions become organized to [Link] are studyingthisprocess using a combination of genetic,
In particular, we are exploiting transposonmolecular and physiological approaches in Antirrhinum.
induced mutations in genes controlling early switches in floral development. These mutations cause
homeotic and heterochronicphenotypes and three categories of genes have been identified. The first
to a floralstate. This gene has
includesfloricaula(lo), which is required to switchinflorescencemeristenms
been isolated and shown to be expressed transientlyin bract, sepal, petal and carpel primordia. The
second group of genes controls the identity (and sometimes the number) of organs in a whorl. These
genes affectoverlapping whorls and their mutant phenotypes suggest a combinatorial model for gene
action in determining the fate of floral primordia. Some of the regulatory interactions between these
genes have been revealed by studyingcis- or trans-actingmutations which have resulted in ectopic gene
[Link] of the thirdcategorydeterminethe identityof organs within one whorl and thus affect
the symmetryof the [Link] propose that the interactionsof these homeotic genes control the basic
but also in a diverse range of
patterns of inflorescenceand flowerdevelopment not only in Antirrhinum,
plant species.

1. INTRODUCTION
Plants develop by the sequential addition of structures
to the main body. These structures originate from
lateral meristemsinitiated on the flanks of the main
apical meristem. The apical meristem is an undifferentiated tissue and throughout developrnent it
integratessignals to determine the identityand positioning of these lateral meristems. Typically, plants
initially undergo a vegetative phase of development
whereby leaves and shoots originate from these meristems. At a later stage, floral induction may occur
whereby these lateral meristemsare induced to form
completely differentstructures,flowers rather than
leafy shoots.
A number of genes affectingthe identity of these
differentineristemshave now been isolated from a
variety of plant species (Coen & Meyerowitz 1991).
Our own study focuses on three classes of gene in
that upon mutation give homeotic and, in
Antirrhinum
some cases, heterochronic phenotypes. These genes
can affect the switch in the type of meristernsproduced, the identity of primordia arising in floral
meristems, or the switching of a meristermfrom
indeterminate to determinate growth.
majushas many genetical and physioloAntirrhinum
gical attributes that make it amenable to study: its
large flowers are easily emasculated, crossed and
scored, the differentphases of vegetative and reproductive development are easily seen and can be
Mhil. Trans. R. Soc. Lond. B (1993) 339, 193-197
P'rintedin GreatBritain

modified environmentally,and its well-characterized

transposable elements allow genes to be isolated and
clonal sectors of wild-type or mutant tissue to be
studied through development (Carpenter & Coen
1990). This paper will focus on Antirrhinum,
although
reference will be made to the other well-analysed
species Arabidopsis,
allowing the conservation or divergence of molecular mechanisms to be compared
between two taxonomically distant species.
2. DESCRIPTION
DEVELOPMENT
STRATEGY

OF WILD-TYPE PLANT
AND MUTAGENESIS

The initial phase of Antirrhinum

growth involves a
vegetative apical meristem that produces pairs of
leaves opposite to each other at each node in a
decussate phyllotaxis (i.e. each pair is at a right-angle
to the previous). Aftera period of vegetative growth
the apex changes to an inflorescencemeristem,producing smaller leaves (bracts) in a spiral arrangement,
generally with a single bract at each node (figure la).
In the axils of these bracts, floral meristems are
initiated which produce four whorls of organs separated by very short internodes, so that consecutive
whorls appear adjacent to each other.
Antirrhinum
flowers are zygomorphic: they can be
divided into mirror-imagehalves by a single plane.
The flowerconsistsof fourwhorls of organs, numbered
from the outside of the flower; thus whorl 1 is

193

[ 55 ]

(? 1993 The Royal Society and the authors

194

D. Bradley and others

offlowering
Generegulation

this transitionmightbe expected to produce proliferating inflorescence shoots in place of flowers. One
mutant obtained with this phenotype is floricaula(flo)
0 8_
Inflorescence
(Carpenter & Coen 1990). The flo mutant initiates
flowers
vegetative growth and the transitionto inflorescence
4-'o
bracts
Ca1
meristemin a similar manner to wild-type. However,
instead of flowers being produced in the axils of
bracts,indeterminateshoots bearing furtherbracts are
leaves
produced, each shoot having two opposite bracts at
the base followed by a spiral of single bracts (figure2).
stem
_~
Each of these shoots in turn produces furthershoots in
the axils of their bracts and this sequence is perpetuated. The wild-typeflo product is thereforenecessary forthe transitionbetween inflorescenceand floral
meristems so that in its absence the inflorescence
(b)
(c)
programme is continually reiterated. The flo mutant
~ stem axis
*
can be considered to be homeotic because it resultsin
B
aborted stamen
one structure(the flower) being replaced by a homologous structure (indeterminate shoot). This mutant
0
X
~~~stamenS
(
can also be viewed as heterochronic (changing the
relative timingof events) because it resultsin an early
2
petals
developmental programme (inflorescence meristem)
~---sepasI1
bract
being continually reiterated.
majus. (a) Schematic
Figure 1. Floweringin Antirrhinum
Other mutants have been described in Antirrhinum,
of wild-typeAntirrhinum
majusdevelopment.(b)
illustration
such as squamosa, which produce inflorescence-like
Floral diagram showingsepal, petal, stamen and carpel
shoots in the axils of bracts, although many of these
organsarrangedin fourconsecutivewhorlsin the axil of a
shoots eventually produce flowers (Stubbe 1966;
bract. (c) Schematicdrawingof the fourwhorls(1-4) and
Huijser et al. 1992). Interestingly,although plants
threeregions(A, B and C) of a flower.
with stable fioricaulaalleles never produce flowers,
under certain conditions the inflorescence meristem
may eventually terminatein carpel-like structures.A
similar mutant, leafy,has been found in Arabidopsis
outermostand whorl 4 is central (figure1). The whorl
(Haughn & Sommerville 1988; Wiegel et al. 1992).
primordia appear sequentially: whorl 1 first,2 and 3
A second class of mutant has almost the opposite
almost simultaneously and finally whorl 4 (Awasthi
effectof those described above because it promotes the
etal. 1984). Organs within a whorl are upper, ifcloser
conversion of inflorescence to floral meristems in
to the stem (adaxial), or lower if nearer the bract
positions where this would not normally occur. Plants
(abaxial). Whorl 1 consistsof five sepals and whorl 2,
with indeterminateinflorescences,such as Antirrhinum,
the corolla, contains five petals which are fused for
do not produce terminal flowers. Thus, the apical
part of theirlength, but end in five separate lobes. In
inflorescence meristem does not itself undergo the
whorl 3, five stamen primordia are initiated, though
mutant in Antirtransitionto floral. The centroradialis
the uppermost fails to develop fully. Whorl 4 is
rlihnum
produces terminal flowers,suggestingthat the
occupied by two united carpels forminga gynoecium
with a bilocular ovary (figure lb).
wild-typegene product inhibits this apical conversion
To isolate and study homeotic genes in Antirrhinum (Kuckuck & Schick 1930).
has been isolated and
The flo gene of Antirrhinum
we have carried out an extensive transposon mutageextensivelycharacterized (Coen et al. 1990). It pronesis experiment (Carpenter & Coen 1990). Plants
duces a transcript of about 1.6 kb which has the
carrying active transposons were grown at 15?C, a
temperature inducing transposition in Antirrhinum potential to encode a protein whose sequence shows
no extensive homologies with any other proteins in
(Carpenter et al. 1987), and self-pollinated to give
available data-banks, though it contains some motifIs
26 000 progeny,many of which were expected to carry
found in [Link] suggeststhat FLO
recessive mutations in a heterozygous condition. The
homozygous phenotypeswere revealed by self-pollinamay be a transcriptional activator, although other
tion and growing 80000 plants. Over 15 independent
roles cannot be excluded. RNA analysis reveals that
homeotic mutations were obtained from this screen
flo is induced within two days when plants are shifted
and here we present our conclusions derived from a
to daylength conditions that promote [Link] situ
hybridization shows thatflo is expressed from a very
study of some of the genes involved.
early stage in wild-type inflorescences in a very
temporal and spatial sequence. The earliest
specific
OF THE FIRST STEPS
CONTROL
3. GENETIC
expression seen is in bract primordia and is followed
DEVELOPMENT
IN FLORAL
by expressionin sepal, petal and carpel primordia,but
The transitionfrom inflorescenceto floral meristems
no expressionis seen in stamen primordia. Expression
representsthe firststep specific to the floral developin each organ is transientand is not observed in later
mental pathway. Mutants that are unable to carryout
stages of development. Taken together, these results
(a)

apical meristem

Phil. 7'Tans.R. Soc. Lond. B (1993)

[ 56

186

F. Grosveld and others

Regulationof humanglobingeneswitching

be capable of interactingwith any of the genes. Such a

mechanism would also explain why the genes compete
with each other for interaction with the LCR, an
observation which is difficult to reconcile with a
mechanism involving differentgenes interactingwith
differentelements of the LCR at the same time (see
below).
5. DEVELOPMENTAL
REGULATION
OF THE
p-LIKE GLOBIN GENES IN TRANSGENIC
MICE
the developmental regulation of the individual
WN5hen
globin genes in the absence of the LCR was studied in
transgenic mice, the s gene was inactive (Shih el al.
1990), but the y and ) globin genes were expressed in a
developmental specificmanner, albeit at low lexvelsand
dependent oIn the position of integration in the host
genome (Mlagram el al. 1985; Townes el al. 1985;
Chada etal. 1986; Kollias etal. 1986). When the genes
were studied in the context of the complete LCR region
in transgenic mice, it was found that the e gene was
expressed at the embryonic stage only (P. Fraser,
unpublished data) in agreement with the data published forthe E gene linked to part of the LCR (Shih etal.
1990; Raich et al. 1990). The y globin gene like its
murine structural homologue the 3hl gene, is
expressed in the embryonic yolk sac. However in
contrast to 3hl the y gene is expressed in the early
foetal liver and is onl silenced after day 16 of
development. It is not expressed in the adult (Dillon &
Grosveld 1991). Although it has not been directly
testedin transgenicmice an indixvidualS gene linked to
the LCR is expressed prematurely in differentiating
embryonicstem cells, although not at maximal levels.
Like the murine 3 gene, the human 5 gene is expressed
at high levels in the foetal liver and adult (Blom et al.
1989; Enver etal. 1990; Behringeretal. 1990; Grosxeld
et al. 1987). This suggeststhat a large part but not all of

9.5 d

N
~~~
04

_
-~

Hy

"

the developmental regulation of the globin genes is

specified by the regions immediately flanking the
genes. However, it is also clear that the LCR is not
developmentally neutral and that it influences the
expression pattern of the genes (see below).
The simplest model on the basis of these data (see
below) was the proposal that stage-specificsuppressors
are the very important regulators of the individual
globin genes (figure 3; Dillon et al. 1991). According
to this model the LCR can interact with each of the
genes at any stage of development but it interacts
preferentiallywith the e gene in the yolk sac. WNhen
the interactionis negated by stage-specificsuppressors
in the foetal liver the LCR would interactpreferentially
with the y globin genes which in turn are suppressed
in the adult. At least two candidate binding sites for
suppressor factors have been proposed, one in the
upstream region of the promoter of the ? globin gene
(Cao et al. 1989) and one at the distal CAAT box in
the promoter of the y globin genes (Superti Furga et
al. 1988; Mlantovaniet al. 1989). However, this model
does not explain how the later expressing y and 3
genes are kept silent at the early, stages of human
erythroid development. Evidence from natural
mutations and transgenic mouse experiments with
combinations of genes suggest that competition of the
genes for the LCR plays an important role in this
process,in particular forthe y and 3 genes (see below).
6. COMPETITION
IN TRANSGENIC

/t

st

HPFH99

T} n_

GENES

Competition between the genes became apparent

when combinations of genes were used in transgenic
mouse [Link] premature expressionof the 3
gene when it was linked to the LCR and expressed in
transgenic mice, could be abrogated by competition
for the LCR with another globin gene, e.g. y or ox
(Behringeretal. 1990; Enver et al. 1990; Hanscombe et

13.5 d

Ce)

OF THE GLOBIN
MICE

so

xo

-o

HPFH61

-s

4
*~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
~o

Figure 3. SI nuclease mapping analysis of minilocus wild-type yrand Greek HPFH y3 mice during development.
RNA was prepared fromtransgenic9.5 day yolk sacs, 13.5 day foetal lixers and fromblood of adult mice. Tlhe yolk
sacs and foetal livers containing minilocus wty- wNere
analysed directl) aftermicroinjectionwhereas yolk sac, foetal
liver and adult blood containing the Greek HPFH yS constructwere obtained by breeding lines 61 and 99. Control
RNAs were fromK562 and Hul l cells. T he signals for humainy- and S-globin RNA are indicated on the left (Hy.

Hr).

Phil. Tranis.R. 'Soc. LoInd.B i 1993

[ 48 1

196

D. Bradlev and others

of flowerina
Gene renigulalion

mine whorl identity. In a purely spatial model,

concentric fields could be set up in the early floral
meristem, independently of the sequence of primordium initiation. The gene functionsrestrictedto the
defined regions A, B and C could be activated in the
appropriate field and hence specifythe fate of primordia growing out from the various regions of the
meristem. This type of model is formally similar to
models proposed for the control of segmentidentityin
DJrosophila
(Ingham 1988). In the sequential type of
model, which shares some featuresof early models of
flowerdevelopment (Heslop-Harrison 1964), the consecutive growth of the primordia would be essential
for the establishmentof the domains of activity. The
various gene functionswould be activated in a manner
that reflectedthe sequence of primordium initiation.
Some supporting evidence for the latter sequential
model has come from studies on the flo gene of
Antirrhinum.
As described above, fio is expressed
sequentially and appears to be activated in regions
where primordia are being initiated. A furtherfeature
offlo expression is that it occurs transientlyin bract,
sepal, petal and carpel but not stamen primordia.
This pattern suggests that fio not only acts to switch
inflorescenceto floralmeristems,but also interactsin a
sequential manner with the whorl identitygenes that
pattern the floral meristem (Coen et cl. 1990).
6. GENETIC

CONTROL

OF WHORL

NUMBER

Some of the whorl identitymutants also affectwhorl

number. Extreme mutants in genes affectingregion B
often give fewer whorls than wild-type. Mutants in
genes acting in region C usually have an increase in
the number of whorls and a more-or-lessindeterminate growth pattern. These findingsreveal two functions of the whorl identitygenes, namely the control of
organ fate and the control of whorl number. The
extent to which these separate roles are mediated by
the homeotic genes activating particular pathways or
target genes is not fullyunderstood.
7. GENETIC CONTROL OF DIFFERENCES
WITHIN A SINGLE WHORL
In radially symmetric flowers (actinomorphic), all
organs in the same whorl have very similar or
identical morphologies, while in zygomorphic flowers
one or more whorls have organs
(e.g. Antirrhinum),
which are distinct. The cycloidea(cyc) mutations in
Antirrhinum
give flowers with a more radially symmetric appearance than wild-type (Stubbe 1966) and
four such mutations were isolated in our transposon
mutagenesis screen. Extreme cyc mutations give
flowerswith petals all resembling the lowest petal of
wild-type. Therefore,cycis a homeotic mutation since
organs of one type are replaced by another type.
Furthermore, cyc appears to affect several whorls,
since all stamens develop in cyc mutants compared to
the abortion of the upper stamen in wild-type.
Does cyc affectorgans in the same way irrespective
of the whorl they occupy? In wild-type,the abortion
of the uppermost stamen depends upon cyc activity.
Phil. Tr-ans.R. Soc. Lond. B (1993)

[ 58

When stamens are found in whorl 2, in the case of ovu

mutants, the upper stamens are again aborted, suggesting that cycinteracts with primordia in a similar
way, irrespectiveof the whorl they occupy. Thus the
fate of a primordium depends on a combination of
interactionsthat determinewhorl identityand relative
upper or lower position.
8. INFLORESCENCE

EVOLUTION

is among the species with zygomorphic

Antirrhinum
flowers that also have indeterminate inflorescences
and so do not normally terminate with a flower.
Nevertheless,exceptional plants which produce terminal flowershave been found among these species and
they develop terminal flowers with radial symmetry
(Peyritsch 1872). This phenotype has been studied in
and shown to be due to a recessiveallele of
Antirrhinum
the locus centroradialis.A few normal flowers are
produced laterally in this homeotic mutant, but the
apical inflorescence soon produces a radially symmetric terminalflower,with all petals resemblingthat
of the lower petal of wild-type plants, similar to the
conversion seen in cyc. An explanation for the symmetryof this terminalflowercan be found in a polarcoordinate model forzygomorphy(Carpenter & Coen
1990). Floral meristemsin axillary positions are in an
asymmetricalenvironmentwith the main apex above
and a bract below them,subjecting them, therefore,to
a gradient of cyc activity. In contrast, a terminal
floweris already in a radially symmetricenvironment
and so may eitherexperience no cycgradient, or rlo cyc
activity itselfmay occur in such an environment.
also have implicaMutations such as centroradialis
tions for inflorescence evolution. Some plants may
produce a large single flowerat the end of theirshoots,
while in most species flowers are clustered to form
inflorescences. In plants with determinate inflorescences, the apical meristemmust convert into a floral
meristem, with other flowers found in axillary positions or terminatingside-branches. An indeterminate
inflorescencecan produce a quite differentarray of
flowers,as each shoot only produces flowersin axillary
positions. The isolation of genes such as fio, that
regulate the transition from inflorescence to floral
meristems,will allow us to address the question of
inflorescence evolution at the molecular level.
Furthermore, the interactions of fo and the organ
identity genes should add insight as to how plants
overcome the constraintsimposed by meristemswhen
tryingto pattern and control sequential organ development.

REFERENCES
Awasthi, D.K., Kumar, V. & Murty, Y.S. 1984 Flower
development in Antirrhinum
majus L.(Scrophulariaceae)with
a comment upon corolla tube [Link]. Mag. Tokyo
97, 13-22.
Bowman, J.L., Smyth, D.R. & Meyerowitz, E.M. 1989
Genes directingflowerdevelopment in [Link]. Cell
1, 37-52.
Carpenter, R. & Coen E.S.
1990 Floral homeotic

Generegulation
offlowering D. Bradley and others 197
mutations produced by transposon-mutagenesisin Antirrhinummajus. GenesDev. 4, 1483-1493.
Carpenter. R., Martin, C.R. & Coen, E.S. 1987 Comparison of genetic behaviour of the transposable element
ITam3 at two unlinked pigment loci in Antirrhinum
majus.
Molec. Gen. Genet.207, 82-89.
Coen, E.S., Romero, J.M., Doyle, S., Elliott, R., Murphy,
G. & Carpenter, R. 1990 floricaula: a homeotic gene
required for flowerdevelopment in Antirrhinum
majus. Cell
63, 1311--1322.
Coen, E.S. & Meyerowitz, E.M. 1991 The war of the
whorls: genetic interactions controlling flower development. Nature,Lond. 353, 31-37.
Drews, G.N., Bowman, J.L. & Meyerowitz, E.M. 1991
Negative regulation of the Arabidopsis homeotic gene
agamousby the apetala-2 product. Cell 65, 991-1002.
Haughn, G.W. & Sommerville, C.R. 1988 Genetic control
of morphogenesis in [Link] Genet.9, 73-89.
Heslop-Harrison, J. 1964 Sex expression in flowering
plants. Brookhaven
[Link]. 16, 109-125.
Huijser, P., Klein, J., Lonnig, W.-E., Meijer, H., Saedler,
H. & Sommer, H. 1992 Bracteomania,an inflorescence

Phil. Trans. R. Soc. Lond. B (1993)

[ 59

anomoly, is caused by the loss of function of the MADSbox gene squamosa in Antirrhinum
majus. EMBO J. 11,
1239-1249.
Ingham, P.W. 1988 The molecular genetics of embryonic
pattern formationin [Link],Lond. 335, 25-34.
Kuckuck, H. & Schick, R. 1930 Die Erbfaktoren bei
Antirrhinummajus und ihre Bezeichnung. Z. indukt.
Abstamm.u. VererbLehre.
56, 51-83.
Peyritsch, J. 1872 Uber Pelorienbildungen. Sber. Akad.
Wiss. Wien. 66, 1-35.
Schwarz-Sommer, Z., Huijser, P., Nacken, W., Saedler, H.
& Sommer, H. 1990 Genetic control of flower development: homeotic genes in Antirrhinum
majus. Science,Wash.
250, 931-936.
Sommer, H., Beltran,J., Huijser, P., Pape, H., Lonnig, W.,
Saedler, H. & Schwarz-Sommer, Z. 1990 Deficiens,a
homeotic gene involved in the control of flower morphogenesis in Antirrhinum
majus: the protein shows homology
to transcriptionfactors. EMBO J. 9, 605-613.
L. sect
Stubbe, H. 1966 GenetikundZytologievonAntirrhinum
Antirrhinum.
VEB. Gustav Fischer Verlag, Jena.