Download Leaf and Flower Development in Pea (Pisum

Annals of Botany 88: 225±234, 2001
doi:10.1006/anbo.2001.1448, available online at http://www.idealibrary.com on
Leaf and Flower Development in Pea (Pisum sativum L.): Mutants cochleata and unifoliata
J E N N I F E R L . YA X L E Y{, W I E SL AW J A B LO N SK I{ and J A M E S B . R E I D *{
{School of Plant Science, University of Tasmania, GPO Box 252-55, Hobart, 7001, Australia and {Central Science
Laboratory, University of Tasmania, GPO Box 252-55, Hobart, 7001, Australia
Received: 23 September 2000 Returned for revision: 3 November 2000 Accepted: 6 April 2001
The stipule mutant cochleata (coch) and the simple-leaf mutant unifoliata (uni) are utilized to increase understanding
of the control of compound leaf and ¯ower development in pea. The phenotype of the coch mutant, which a€ects the
basal stipules of the pea leaf, is described in detail. Mutant coch ¯owers have supernumerary organs, abnormal fusing
of ¯ower parts, mosaic organs and partial male and female sterility. The wild-type Coch gene is shown to have a role
in in¯orescence development, ¯oral organ identity and in the positioning of leaf parts. Changes in meristem size may
be related to changes in leaf morphology. In the coch mutant, stipule primordia are small and their development is
retarded in comparison with that of the ®rst lea¯et primordia. The diameter of the shoot apical meristem of the uni
mutant is approx. 25 % less than that of its wild-type siblings. This is the ®rst time that a signi®cant di€erence in
apical meristem size has been observed in a pea leaf mutant. Genetic controls in the basal part of the leaf are
illustrated by interactions between coch and other mutants. The mutant coch gene is shown to change stipules into a
more `compound leaf-like' identity which is not a€ected by the stipules reduced mutation. The interaction of coch and
tendril-less(tl) genes reveals that the expression of the wild-type Tl gene is reduced at the base of the leaf, supporting
# 2001 Annals of Botany Company
the theories of gradients of gene action.
Key words: Pisum sativum, garden pea, leaf morphogenesis, compound leaf, leaf mutants, ¯ower morphology.
I N T RO D U C T I O N
The garden pea (Pisum sativum L.), along with tomato
(Lycopersicon esculentum Miller), is a model species for
understanding compound leaf development. The pea leaf is
pinnately compound and consists of basal, foliaceous
stipules, proximal lea¯ets and distal tendrils (Fig. 1A). In a
major review of pea leaf development mutants, Marx (1987)
suggested that various mutants appear to operate in these
three di€erent `domains' within the leaf (basal, proximal and
distal), and that the action of these leaf development genes
might help to de®ne these areas. However, studies of pea leaf
development have concentrated on the recessive mutant
a®la(af) (Kujala, 1953; Goldenberg, 1965), the semidominant mutant tendril-less(tl) (Vilmorin and Bateson,
1911) and the af tl double mutant (Fig. 1D, F and G), which
alter lea¯et and tendril pinna identity in the proximal and
distal domains (Meicenheimer et al., 1983; Gould et al.,
1986, 1994; Cote et al., 1992; Lu et al., 1996; Villani and
DeMason, 1997, 1999a, b, 2000). The stipules and basal
compartment have received little attention, with only one
model of pea leaf development (Gourlay et al., 2000)
considering the cochleata (coch) mutant (Wellensiek, 1959),
in which the stipules of the middle nodes of the plant are
compound (Fig. 1B).
This study gives a detailed description of the phenotype
of the coch mutant and investigates some of its interactions
with other pea leaf genes to increase knowledge of genes
controlling the basal domain of the pea leaf. The
* For correspondence. Fax 61 3 62 26 2698, e-mail Jim.Reid@utas.
edu.au
0305-7364/01/080225+10 $35.00/00
interactions of coch with tl in the basal compartment and
of coch with the recessive mutant stipules reduced (st) are
described. As the names suggest, st mutants have stipules
that are reduced to small, narrow structures (Pellew and
Sverdrup, 1923) (Fig. 1C) and the leaves of tl mutants have
lea¯ets in place of terminal tendrils (Vilmorin and Bateson,
1911) (Fig. 1D). Interactions between these mutants provide information about domains of gene activity that are
not evident in the single mutants.
Meristem size has been linked to the size of the structures
they produce (Smith and Hake, 1992; Sundberg and Orr,
1996; Lijsebettens and Clarke, 1998; Jackson and Hake,
1999). It has been suggested that di€erences in meristem or
primordial size could produce di€erent pea leaf mutants
(Young, 1983; Gould et al., 1992), although no size
di€erence was found in the meristems of af, tl and af tl
mutants (Meicenheimer et al., 1983; Gould et al., 1986;
Cote et al., 1992). Here, we examine the early leaf
development of coch and unifoliata (uni) (Eriksson, 1929)
mutants to investigate the role of meristem size in the
development of di€erent types of leaf structures in addition
to lea¯ets and tendrils. Strong alleles of the uni mutant
change the multiple lea¯ets and tendrils of a wild-type leaf
into a single leaf blade with basal stipules (Lamprecht,
1933) (Fig. 1E) and seem to merge the proximal and distal
compartments into one. The uni gene is the pea homologue
of the ¯oral meristem identity genes LEAFY in arabidopsis
and FLORICAULA in Antirrhinum (Hofer et al., 1997). Pea
shoot apical meristems were examined by environmental
scanning electron microscopy (ESEM), since this allows
# 2001 Annals of Botany Company
226
Yaxley et al.ÐLeaf and Flower Development in Pea
here was generated by J. Weller in Hobart through
mutagenesis of the cultivar Torsdag (Hobart line 107).
Seeds were treated with 1 % EMS (ethylmethanesulfonate)
for 6 h at 18 8C. Among 1100 M2 families, one showed the
leaf-like stipules typical of coch mutants. This mutant line
(AF99) was crossed to a coch type line, JI 2757, produced
by S. Blixt (Blixt, 1972; WL5446) from the variety Parvus
and provided by M. Ambrose (John Innes Institute, UK).
The F1 plants of this cross had leaves typical of coch plants,
including leaf-like stipules in the middle nodes, indicating
that our coch mutant line AF99 possessed a mutant allele at
the known cochleata locus. AF99 has been backcrossed
twice with its progenitor line 107.
To generate the coch st double mutant, coch (AF99)
(male parent) was crossed with Hobart multiple marker line
HL31 ( female parent) which is homozygous recessive for st.
To examine the expression of Tl in a coch background, coch
(AF99) (male parent) was crossed with Hobart multiple
marker line HL111 ( female parent) which is homozygous
recessive for the leaf development genes tl, af and st.
Identi®cation of coch st double mutant plants
F I G . 1. The wild-type pea leaf (A) with basal stipules (S), proximal
lea¯ets (L) and distal tendrils (T) and leaves of various leaf mutants: B,
cochleata(coch); C, stipules reduced(st); D, tendril-less(tl); E, unifoliata(uni); F, a®la(af); and G, the af tl double mutant. Bars ˆ 4 cm.
objects to be seen in their natural state (Danilatos, 1993;
Jablonski, 1997).
Two wild-type F1 heterozygotes from the cross of coch
(AF99) (male parent) and st (HL31) ( female parent) were
grown; from these plants 40 F2 seeds were sown. It was
determined that the cross was successful because the F1
plants had wild-type stipules. However, the identity of the
double mutant plants amongst the 40 plants of the F2
generation was not immediately obvious, as both coch and
st mutations have the e€ect of reducing stipules. The coch st
double mutants were identi®ed from their ¯oral
morphology, and their leaf phenotype was then examined.
(Homozygous coch plants have mutant ¯owers while the st
mutation does not a€ect ¯owers.) The F3 progeny of seven
F2 plants with reduced stipules (st st) and wild-type ¯owers
were grown. Two-thirds of these plants were expected to be
heterozygous for the coch mutation (Coch coch); however,
only one of the seven F3 families planted showed
segregation of ¯ower phenotype. In this family there were
ten plants with reduced stipules and wild-type ¯owers
(Coch±st st), and also two plants with both reduced stipules
and compound stipules and mutant coch-like ¯owers
( putative coch coch st st plants). The F4 progeny from the
putative double mutant plants (14 plants), and some of
their heterozygous (Coch coch st st) siblings (16 plants) were
grown to con®rm the double mutant phenotype, which was
clear based on ¯ower morphology in the homozygous st
background.
M AT E R I A L S A N D M E T H O D S
Plant material
Identi®cation of coch coch Tl tl plants
The lines of Pisum sativum L. used during this work are
held in the collection at Hobart, Australia. The uni line JI
2171 ( provided by J. Hofer, John Innes Institute, UK), is
descended from the original spontaneous mutant described
by Eriksson (1929). This line is maintained as a heterozygote, and in our meristem studies uni plants were
compared with their wild-type siblings. The coch line used
Four F1 heterozygotes from the cross of coch (AF99)
(male parent) and HL111 ( female parent) were grown, and
from these plants 64 F2 seeds were sown (16 from each F1
plant). The cross was found to be successful as the F1 plants
had wild-type leaves. Leaf forms of the ten F2 plants which
were heterozygous for Tl (Tl tl) and had coch leaf-like
stipules were observed. Plants heterozygous for Tl can be
Yaxley et al.ÐLeaf and Flower Development in Pea
distinguished because their leaves have ¯attened tendrils
(White, 1917; Villani and DeMason, 1999b).
Growing conditions
All plants were grown in a 1 : 1 mixture (v/v) of
vermiculite and 10 mm dolerite chips topped with 2±3 cm
of pasteurized potting mix (1 : 1 mixture (v/v) of coarse river
sand and peat moss, with added macro and micro
nutrients). Two plants were grown in each 14 cm slimline
pot. Before planting the seeds, a small section (about
3 mm2) of testa was removed with a sharp razor blade to
facilitate even germination. Seeds were then coated in
Thiram 800 fungicide (active ingredient Thiram 800 g kg ÿ1;
Agchem, Para®eld Gardens, SA, Australia) to help prevent
fungal infection during germination, and planted into wet
soil. The pots were watered sparingly for the ®rst 5 d until
seedlings began to emerge. Watering was then carried out
three times a week for 3 weeks, and then daily thereafter.
Plants were provided with nutrient solution [1 g l ÿ1
Aquasol (Hortico, Laverton North, Australia) and 0.05 g
l ÿ1 iron chelate (Kendon Chemicals, Thornbury, Australia)]
weekly. Vertical strings were used to support the plants
once they reached 10 cm in height. The photoperiod was
18 h, consisting of natural daylight extended by mixed
incandescent (100 W bulbs) and ¯uorescent (40 W cool
white tubes) light, giving approx. 25 mmol m ÿ2 s ÿ1 at the
pot surface. Growth of the segregating populations and of
plants used to examine in¯orescence structure took place
during autumn/winter with mean maximum/minimum day
and night temperatures of approx. 22 and 14 8C, respectively. Plants used for microscopic examination and for all
other morphological measurements were grown during
spring, with mean maximum/minimum day and night
temperatures of approx. 25 and 15 8C, respectively. Plants
were treated as required with fungicide or pesticide.
Scoring of morphological characters
Counting of nodes was acropetal, with the cotyledons as
node 0, the two subsequent scale leaves as nodes one and
two and the ®rst true leaf as node three. Leaves at the apex
were numbered using the plastochron index (Lamoreaux
et al., 1978; Sylvester et al., 1996) to describe their state of
development. The plastochron one (P1) leaf is the youngest
leaf which has been initiated on the ¯ank of the apex, the
plastochron two leaf (P2) is the next oldest primordium, and
so on. Plastochron zero (P0) marks the site where the next
leaf will be initiated on the ¯anks of the apical dome.
Means were compared using Student's t-test.
Environmental scanning electron microscopy
Three mutant and three wild-type leaves, for both the
coch and uni mutants and their corresponding wild types,
were examined for each of nodes 8 to 12. To observe each
successive node, groups of plants were harvested every 2±
3 d. Ten plants of each mutant line and the corresponding
wild type were examined 4 weeks later to determine the
227
typical mature leaf morphology of each node examined
during development.
The whole apices of plants were harvested directly before
they were to be examined. The meristem was then dissected
from the apical leaves under a stereo-microscope with a
minimum light intensity and with the sample surrounded by
a pool of water to prevent dehydration. The apex was then
attached to an aluminium stub with double-adhesive
conductive tape and placed on a Peltier-e€ect thermoelectric cold stage (Omega CN 900A, ElectroScan, Wilmington,
Australia) attached to an environmental scanning electron
microscope (ESEM 2020, ElectroScan). The specimen
chamber was closed and the temperature reduced to 2.5±
2.8 8C before the pressure was reduced. The specimen was
localized and the primary focus and video signal level
established at a water vapour pressure of 8 Torr. The
pressure was then slowly reduced to 6.0±7.5 Torr (to
minimize water on the specimen surface). Once an image
was acquired (at 2048 2048 resolution), Ilford FP±4 (125)
®lm was used to obtain a hard copy.
Using this technique the apices do not require any
treatment before microscopy, thus substantially reducing
the possibility of artefacts (Crang and Klomparens, 1988).
The structural integrity of the specimen is maintained when
the internal pressure corresponds to the saturated water
pressure at a given temperature. (The surrounding medium
also acts as a neutralizing agent to prevent primary electron
charge build-up on the surface of the specimen.) Measurements of apical meristem size were made on screen during
ESEM observation.
R E S ULT S
Phenotype of the coch mutant
The coch mutation a€ects only the stipules of the leaf;
lea¯ets and tendrils are unchanged. Stipules were mostly
absent from the ®rst few nodes (nodes 3-5). Above this
(nodes 6 and 7), coch stipules were sometimes small and
straplike, sometimes sessile and elliptical, or sometimes
spatulate. At higher nodes (nodes 8-11), the stipules were
largely compound, with proximal lea¯ets and distal tendrils,
that re-iterated the structure of the leaf blade (Fig. 1B).
Above node 11, leaf complexity was reduced, and spatulate
and sessile-elliptical stipule forms became common;
occasionally thread-like forms occurred (Figs 2 and 3).
Approximately 25 % of coch leaves had mismatched
stipule pairs. For example, an elliptical sessile stipule
occurring opposite a spatulate stipule (e.g. node 12,
Fig. 2), or a compound stipule opposite a thread-like
stipule. In addition, approximately 10 % of compound
coch stipules had lea¯et and tendril pairs which were not
opposite each other on the stipule rachis (e.g. node 11,
Fig. 2). This occurred less often (approximately 2 %) on the
main rachis of wild-type leaves.
The stipules of coch mutants did not have visible axillary
buds. This applied to both compound coch stipules and
simple coch stipules. Thus, coch compound stipules appear
to re-iterate only the proximal and distal parts of the
compound pea leaf (the lea¯ets and tendrils). They do not
228
Yaxley et al.ÐLeaf and Flower Development in Pea
F I G . 2. The stipule morphology of a single coch mutant. The pairs of
stipules from node 5 (top right) to node 14 (bottom left) are shown.
Percentage of stipule forms
have stipules or axillary buds at their bases, as occurs in true
compound leaves. In addition, almost half of the coch plants
with compound stipules had no apparent axillary bud in the
main leaf axil. This compares with an approximately 3 %
occurrence of `blind' leaf axils in wild-type leaves.
The node at which leaves with two pairs of lea¯ets ®rst
appeared was similar for coch (11.3 + 0.2) and wild-type
(11.8 + 0.2) plants (0.1 4 P 4 0.05, n ˆ 24). On average,
compound coch stipules also changed to two lea¯et pairs at
the same node (10.4 + 0.7) as the main leaf rachis
(11.3 + 0.2) (0.4 4 P 4 0.2, n ˆ 24). However, the compound stipules exhibited more variability, with some
changing to two lea¯et pairs substantially earlier (node 7)
than the main leaf rachis, and other compound coch stipules
never showing more than three lea¯ets (one lea¯et pair and
a mixed lea¯et-tendril pair). Flowering time and ¯owering
node were also not signi®cantly a€ected. Mutant coch
plants ¯owered at node 15.5 + 0.9 in 65.2 + 0.9 d, and
wild-type plants ¯owered at node 16.0 + 0.3 in 65.4 + 0.6 d
(0.8 4 P 4 0.7 for ¯owering node; 0.9 4 P 4 0.8 for
¯owering time, n ˆ 24).
Wild-type pea ¯owers have ®ve sepals, two fused keel
petals, two wing petals and a standard petal ( ®ve petals in
all), ten anthers (nine fused into a ®lament tube and one
partially free) and a single central carpel (Tucker, 1989;
Ferrandiz et al., 1999) (Figs 4 and 5). Flowers of coch
mutants ranged from nearly normal in appearance (Fig. 4),
to open ¯owers with supernumerary organs in each whorl,
abnormal organ fusing and some organs which were a
mosaic of di€erent organ types (Fig. 5, Table 1). The more
severely a€ected ¯owers occurred at later nodes. The ®rst
formed ¯owers on the main stem and on laterals were the
most normal in appearance and set the most seed. Mutant
coch ¯owers were largely self-sterile and coch plants
normally produced no more than ten seeds per plant
(compared with an average of 60 seeds per plant for the
wild-type progenitor line). Flowers of coch mutants were
partially sterile when cross-fertilized with wild-type pollen,
and the pollen from coch mutants also showed reduced
fertility when it was used to pollinate wild-type ¯owers. Due
to the low fertility of coch plants, they produced more
lateral branches than the wild type.
There was an increase in sepal number in coch ¯owers
(Table 1) and mosaic sepal/petal organs appeared in the
sepal whorl in ¯owers with severe mutant phenotypes. There
was also an increase in the number of wing and keel petals
(Table 1), and the keel petals did not pair and fuse properly
to make a keel (Figs 4 and 5). Based on petal shape and
colour (white or green), some petals appeared to be a mosaic
of wing and keel petals. In ¯owers with a weak mutant
phenotype, there was no change in the number of standard
petals, but in ¯owers with a strong mutant phenotype two or
three standard petals were common (Table 1), forming a
radially symmetric ¯ower structure (Fig. 5). All coch petals
were narrower at the base (Fig. 6) than those of the wild
type. Wild-type petals form an enclosing sheath around the
100
STIPULE
FORM
75
absent
elliptical
50
spatulate
strap-like
25
0
compound
threadlike
3
4
5
nodes formed
in the embryo
6
7
8
9 10 11 12 13 14 15 16 17
Node
average node
of first flower
F I G . 3. Occurrence of di€erent stipule forms at di€erent nodes of mutant cochleata plants (n ˆ 40).
Yaxley et al.ÐLeaf and Flower Development in Pea
229
T A B L E 1. Number of ¯oral organs in wild-type and
cochleata mutant ¯owers
Average number of organs + s.e.
Floral organ
6.0 + 0.2b
7.1 + 0.4c
Anthers
1.0 + 0a
2.0 + 0a
2.0 + 0a
10.0 + 0a
1.0 + 0a
2.1 + 0.1a
2.8 + 0.1b
11.8 + 0.4b
2.7 + 0.2b
2.9 + 0.3b
2.5 + 0.2b
12.6 + 0.8b
Carpels
1.0 + 0a
1.0 + 0a
1.8 + 0.6b
Petals
Standard
Wing
Keel
F I G . 5. coch ¯owers with strong mutant phenotypes (a wild-type ¯ower
is shown on the far left). In addition to the changes seen in Fig. 4, more
severe mutant ¯owers may have two (A, B) or three (C) standard petals
and two to four gynoecia (E and F: these ¯owers are several days old
which has allowed the carpels to grow out and become more visible).
Flower A has an additional, less developed, ¯ower in the panicle, and
¯ower B has another small ¯oral structure arising from its pedicel.
coch mutant weak coch mutant strong
phenotype
phenotype
(n ˆ 20)
(n ˆ 20)
5.0 + 0a
Sepals
F I G . 4. coch ¯owers with weak mutant phenotypes (a wild-type ¯ower
is shown on the far left). The mutant ¯owers with weak phenotypes
have additional wing and keel petals which become disorganized, and
additional sepals and anthers.
Wild-type
(n ˆ 50)
Means with di€erent letters in the same row are signi®cantly di€erent
(P 5 0.02).
anthers and stigma, whereas the narrower petal bases of the
coch ¯ower cause the petals to fall open so the anthers and
stigma are not enclosed within the keel. This means that coch
¯owers have a more open structure. Their pollen dried
quickly and tended not be deposited on the stigma(s), thus
contributing to their reduced fertility.
The number of anthers was also altered in coch ¯owers
(Table 1). Occasionally less than ten anthers were present,
but generally ¯owers had more than ten anthers (up to 17 in
¯owers with severe mutant phenotypes). Anthers of coch
¯owers were commonly fused together above the level of the
®lament tube, and also fused to petals above the base (the
outer stamens are normally basally adnate to the petals;
Tucker, 1989). In coch mutant ¯owers with severe
phenotypes, most of the anthers were abnormally fused.
The gynoecium remained unchanged in coch ¯owers with
weak mutant phenotypes, whereas in highly disturbed
¯owers up to four carpels were present (Fig. 5, Table 1).
Wild-type pea ¯owers normally have two ¯owers per
panicle. This varies somewhat at di€erent nodes, with more
single-¯owered panicles at the highest nodes (Hole and
Hardwick, 1976; Fig. 7). The number of single-¯owered
panicles also varied with planting season; more single¯owered panicles occurred during winter than spring (data
not shown). In mutant coch plants, there was a greater
occurrence of single-¯owered panicles at all nodes at both
planting dates (Fig. 7). This varied with node and season in
a similar manner to the wild type, with more single-¯owered
panicles occurring at higher nodes and during winter.
Approximately 10 % of coch panicles contained one ¯ower
plus a partially developed ¯ower or ¯ower bud (Fig. 5). In
addition, approximately 10 % of coch pods have a `collar' of
leafy bract tissue with crenulate edges surrounding the base
of the pedicel.
Interaction of coch and st
F I G . 6. Mutant coch ¯owers (left) have narrowed petal bases compared
with the wild type (right). Whole ¯owers are shown at the top and a
single standard petal at the bottom. (Petal bases are narrower in coch
mutant ¯owers even when there are no extra petals in the whorl.)
The st mutant had stipules reduced to small strap-like
structures (Fig. 1C). This morphology is the same at each
node, and other leaf parts and ¯owers are normal (Pellew
and Sverdrup, 1923). The coch st double mutant had
230
Yaxley et al.ÐLeaf and Flower Development in Pea
Single flowers (%)
100
80
60
coch
wild-type
40
16
18
Node
20
22
F I G . 7. Occurrence of single-¯owered panicles in mutant coch plants
and their wild-type progenitor (`Torsdag') for a winter planting.
(Panicles of coch mutants with one ¯ower plus a partially developed
bud were considered to be panicles with more than one ¯ower.) n ˆ 13.
coch-like stipules, indicating that coch is largely epistatic to
st in this background. The coch st double mutants showed all
the stipule forms typical of coch plants: compound, elliptical
and sessile, strap-like, spatulate and ®lamentous. However,
the double mutant plants had signi®cantly fewer
(P 5 0.001) compound stipules than coch plants. Mutant
coch plants had 7 + 0.3 nodes carrying compound stipules
and coch st double mutant plants had 4 + 0.5 nodes with
compound stipules (n ˆ 24).
Interaction of coch and tl
Leaves of the tl mutant have lea¯ets in distal positions
where tendrils are normally present (Fig. 1D). The tl
mutation is incompletely dominant, and heterozygotes
can be distinguished by the presence of ¯attened, rather
than cylindrical, tendrils. Compound coch stipules showed
reduced expression of the wild-type Tl gene compared with
the main leaf rachis. Leaves of the genotype coch coch Tl tl
had ¯attened tendrils typical of Tl tl heterozygotes in the
distal region of the leaf, but the compound stipules had
only lea¯ets (Fig. 8), like the leaf blades of the homozygous
recessive tl mutant.
A similar reduction in Tl expression was seen in the
compound stipules of coch coch af af Tl tl plants. The leaf
blade of coch coch af af Tl tl plants had tendrils like that of
an af mutant leaf, with the tendrils somewhat ¯attened at
their tips due to the incomplete dominance of the Tl gene.
However, the compound stipules of this genotype consisted
of branched tendrils ending in small terminal lea¯ets (Fig. 9),
similar to the af tl double mutant phenotype (Fig. 1G).
Early leaf development of coch and uni
In the coch apical meristems examined, stipule primordia
were smaller and less developed than the primordia of the
®rst lea¯et pair, from the time they were ®rst visible
(Fig. 10B) until late stages of leaf expansion (e.g. P9 ,
Fig. 11). In contrast, wild-type leaf primordia have early
stipule primordia (P2 to P4) that are larger than the primordia
F I G . 8. Leaf-like stipules of the genotype coch coch Tl tl have small
terminal lea¯ets typical of the homozygous recessive genotype tl tl
(arrow). The main leaf rachis has ¯attened tendrils typical of a Tl tl
heterozygote. Bar ˆ 4 cm.
F I G . 9. Leaf-like stipules of the genotype coch coch af af Tl tl have
small terminal lea¯ets typical of the double recessive genotype af af tl tl
(arrow). The main leaf rachis has the ¯attened tendrils typical of a Tl tl
heterozygote. Bar ˆ 4 cm.
of the ®rst lea¯et pair (compare Fig. 10A and B). During P5 ,
compound coch stipule primordia initiated lateral pinna.
During P6 these ®rst stipule lateral pinna became more
dorsiventral and secondary lateral pinna were initiated
(Fig. 12). The diameter of the apical dome of coch mutant
plants, when P0 is node 11, was similar (190 + 4.7 mm) to
that of the corresponding wild type (200 + 5.1 mm) (n ˆ 6).
The phenotype of the uni mutant has been described
previously (Eriksson, 1929; Lamprecht, 1933; Hofer and
Ellis, 1996; Hofer et al., 1997). Flowers of the uni mutant
have an incomplete sepal whorl, no petals or stamens, and
an open gynoecium, with numerous iterations of axillary
¯owers (Hofer et al., 1997). At the nodes examined (7-13),
all uni leaves consisted of normal stipules with a single leaf
blade. Approximately half of these uni leaves were lobed
(Fig. 13), more commonly at the lower nodes examined
(nodes 7±9: 70 % lobed leaves). Leaves of uni plants at the
higher nodes examined (nodes 11-13) had longer petioles
than those leaves at nodes 7±9 which were mostly sessile
(80 % sessile or almost sessile; Fig. 13).
Leaf primordia of the uni mutant initiated a single large
lea¯et primordium and two normal stipule primordia
(Fig. 10C). At the same stage, the wild type had smaller
lea¯et primordia with two large lateral stipule primordia
(Fig. 10A). During P3 , the uni leaf primordium developed a
groove on its adaxial side, indicating the incipient midrib.
This was more obvious in the P4 uni lea¯et (Fig. 10C), which
Yaxley et al.ÐLeaf and Flower Development in Pea
231
development between wild type, coch and uni plants are also
seen when the position in which hairs ®rst appear on these
leaves is compared; in the wild type they appear on stipules in
P3 ±P4 (Fig. 10A), in coch they appear on the ®rst lea¯et pair in
P3 ±P4 (Fig. 10B) and in uni they ®rst appear on the leaf blade
and stipule primordia in P3 ±P4 (Fig. 10C).
The diameter of the apical dome of uni plants, when P0 is
node 11, was signi®cantly less (148.5 + 3.3 mm) than that of
the corresponding wild-type siblings (193.6 + 3.6 mm)
(P 5 0.001, n ˆ 7) (see Fig. 10). However, the node and
time of ®rst ¯ower initiation were similar in mutants and
wild types. Mutant uni plants ¯owered at node 22.6 + 0.4 in
58.9 + 0.4 d in summer, and their wild-type siblings
¯owered at node 21.6 + 0.3 in 57.9 + 0.4 d (both
0.1 4 P 4 0.05, n ˆ 20).
DISCUSSION
Phenotype of the coch mutant
F I G . 10. A, A wild-type apex from line JI2171 with the largest leaf
subtending node 12. The stipule primordia of the P3 and P4 leaves (S3
and S4) are larger and more developed than the ®rst lea¯et primordia
(L3 and L4). The L4 lea¯et primordia are just beginning to develop an
adaxial groove. B, A coch apex from line AF99 with the largest leaf
subtending node 12. An in¯orescence meristem is also present (IF). The
stipule primordia of the P3 and P4 leaves (S3 and S4) are small and
undeveloped compared with the primordia of the ®rst lea¯et pairs (L3
and L4). The stipule primordia are also small and undeveloped
compared with the wild-type stipule primordia (see Fig. 11B). C, The
apex of a uni mutant from line JI2171, with the largest leaf (P4)
subtending node 12. The P3 primordium shows a single large leaf blade
primordium and stipule primordia (S3). The P3 leaf blade primordium
has a central adaxial groove where the midrib is forming. The P4 leaf
blade primordium shows a clear central channel and is developing a
laterally ¯attened folded lea¯et shape. The P3 and P4 leaf blade
primordia also show the development of lobes on their ¯anks (lo).
was clearly a laterally ¯attened structure, with two halves
folded along a central groove. The wild type lea¯et primordia
showed no lateral ¯attening or midrib-groove development
until late P4 (Fig. 10A). The di€erences in early leaf
The stipule forms displayed by the coch mutant vary with
node of insertion. The ®rst ®ve to six nodes of a pea plant are
formed in the developing embryo (Gould et al., 1987; Villani
and DeMason, 1997). Nodes 3-5 usually have no stipules in
the coch mutant, so there may be an interaction between the
environment of the embryo and the coch mutation which
results in the absence of stipules, or the formation of only
small stipules. The vegetative nodes of the coch plant which
develop after germination show predominantly compound
stipule forms. After ¯owering commences (nodes 15-16), the
stipule forms found in coch mutants are simpli®ed. Again, it
seems likely that the major change in the condition of the
shoot apex is responsible for the corresponding change in
stipule form, which parallels the increase in leaf blade
complexity in the middle nodes, and the decrease in leaf
complexity after ¯owering, seen in wild-type plants
(Wiltshire et al., 1994; Villani and DeMason, 2000).
In experiments in which excised pea leaf primordia were
grown in culture (Gould et al., 1994), it has been shown that
the two pinnae of a lea¯et or tendril pair are not always
determined simultaneously. The tendency for opposite
pinnae to develop into di€erent structures is increased in
coch compound stipules. Since the development of coch
stipules is retarded (Gourlay et al., 2000; Figs 10 and 11), it is
possible that they remain meristematic for longer, which
could lead to their greater developmental ¯exibility. In
addition, it was more common for the pinnae of coch
compound stipules not to be positioned opposite each other.
This is a shoot-like characteristic (Sattler and Rutishauser,
1992; Lacroix and Sattler, 1994), and supports the contention
that the coch mutation reduces determinacy in the stipules.
It seems likely that the open structure of the coch ¯ower
will reduce the viability of pollen and the receptive life of
the stigmatic surface. However, while generating crosses in
this study, we also found poor seed set of coch ¯owers
pollinated by hand with wild-type (Coch) pollen. This is
consistent with structural and developmental defects in the
gynoecium of coch mutants that reduce seed set; such
abnormal development of the gynoecium and of the female
gametophyte in putative coch ¯owers was described by
232
Yaxley et al.ÐLeaf and Flower Development in Pea
F I G . 11. Stipules of coch mutants remain less developed than wild-type stipules at later stages of leaf growth. A, P7 leaves at node 13 of coch
mutant (left) and wild-type (right) plants; B, P8 leaves at node 12 of coch mutant (right) and wild-type (left) plants; C, P9 leaves at node 11 of coch
mutant (left) and wild-type (right) plants. (In A and B the apical meristem of the coch mutant has been removed for clarity.)
F I G . 12. P6 coch stipule primordia (S6), from node 11, showing three
pairs of lateral primordia (l1 , l2 , l3). The primary pair (l1) is becoming
¯attened, indicating that they will become lea¯ets. The P6 lea¯et (L6),
and the P5 leaf are also indicated.
Molhova et al. (1988). In the putative coch line examined by
Molhova there was sometimes no ovule formation and
sometimes early formation of the embryo sac was
disturbed. Molhova et al. (1988) also described the
production of feminized stamens in coch ¯owers, leading
to the production of multipistillate ¯owers. While the
production of multipistillate ¯owers through an increase in
carpel number was observed in this study, no feminization
of the stamens was seen. This di€erence could be due to the
di€erent lines examined. Molhova et al. (1988) found a high
percentage of sterility in pollen from coch mutants. Pollen
from the coch mutant utilized in this study also showed
reduced fertility when crossed onto wild-type ¯owers, thus
it is also likely to show some abnormalities.
It is interesting to note that it is the base of the wing and
keel petals that is altered in all coch ¯owers, even in those
with the weakest phenotypes. This suggests that there may
be some homology between the base of the petals and the
base of the leaves (stipules) in the action of the Cochleata
gene. The defective petal bases may be partly responsible
for the improper fusing of the keel petals and anthers. In
F I G . 13. A lobed, almost sessile uni leaf (node 8).
pea, the petals and stamens originate from a common
primordium (Tucker, 1989; Ferrandiz et al., 1999). Incorrect separation of these organs may result in the fused
stamens and petals seen in coch ¯owers. The formation of
some ¯oral organs which are a mosaic of di€erent organ
types indicates that the coch mutation also has an e€ect on
¯oral organ identity.
The reduced formation of two-¯owered panicles seen in
coch in¯orescences shows that the early development of the
in¯orescence is also altered by the mutation. In pea ¯owers,
once the in¯orescence meristem has formed and the
meristematic region which will form the ¯ower has become
distinct, the remaining region will form another ¯ower
initial, or will revert to a vegetative state and form a small
stub (Hole and Hardwick, 1976). In coch ¯owers, a second
¯oral initial is much less likely to form (there are more
single-¯owered panicles), and in addition, more developmental ¯exibility is seen as half-formed second ¯owers are
sometimes seen in coch mutants (Fig. 5), whereas these
forms do not appear in the wild type. Thus, the Coch gene
has been shown to have pleiotropic e€ects on leaf and ¯ower
development, altering stipule form, ¯oral organ number,
shape and identity, fertility and in¯orescence structure.
Yaxley et al.ÐLeaf and Flower Development in Pea
Interaction of coch and st
Double mutant coch st plants resembled coch plants. This
is in contrast to the ®ndings of Marx (1987) and Gourlay
et al. (2000) who reported that coch st double mutants had
no stipules. This may be due to di€erences in the genetic
background and/or allele-speci®c di€erences. In this study,
coch is predominantly epistatic to st. It is possible that the
coch mutation changes the fate of the stipule cells into more
`leaf-like' cells and thus prevents the action of st, which acts
only on stipule cells. Epistasis occurs at all nodes and all
stipule forms are seen in coch st double mutant plants. The
suggestion that coch alters the fate of stipule cells is strongly
supported by the fact that coch compound stipules behave
genetically as if they were leaf blades in a basal position
(Marx, 1987). For example, the coch af double mutant has
stipules which consist of tendrils like the leaf blade.
Interaction of coch and tl
The interaction between coch and tl suggests there is a
gradient of wild-type Tendril-less gene action in the leaf,
with Tl action being reduced at the base of the leaf. This
cannot normally be seen as Tl does not act on the stipules,
but in the compound stipules of coch the reduced action of
the wild-type Tl gene in the base of the leaf is evident. This
supports proposals that gradients of gene action could be
involved in determining leaf form in pea (Hofer and Ellis,
1996; Lu et al., 1996). Lu et al. (1996) suggested that a
gradient of Tl gene function would be a plausible
explanation of pea leaf morphology, with Tl function
producing a tendril-inducing or branching-inhibiting morphogen at the leaf tip. Villani and DeMason (1999b)
showed, via morphological measurement of a series of
mutants with di€erent numbers of tl alleles, that Tl acts in
both the proximal and distal parts of the leaf, but has a
greater e€ect in the distal region.
Early leaf development of coch and uni mutants
The small size of coch stipules and their slow growth and
development strongly suggests that coch stipules are
developmentally retarded compared with lea¯et primordia
of the same leaf. Early slow growth of coch stipules was also
shown by Gourlay et al. (2000). It is possible that this is the
primary consequence of the coch gene mutation, and that
because of their developmental retardation and probably
more meristematic nature, coch stipules respond inappropriately to normal developmental signals, and so develop
into complex compound structures. The formation of
compound stipules seems to be due, at least in part, to
the expression of Uni seen in coch S3 and S4 stipules, which
is not seen in wild-type stipules (Gourlay et al., 2000) (Uni is
important in maintaining blastozone activity). However,
this ectopic expression of Uni in compound coch stipules is
not seen until after morphological changes have occurred in
coch primordia (at late P1 or early P2), and it is seen only in
the compound coch stipules, not in the other more simple
coch forms; thus, this alteration in Uni expression is not
likely to be the primary e€ect of the coch mutation.
233
The e€ects of the mutations coch and uni seem to be
apparent at an earlier stage (late P1 to early P2) of leaf
primordia development than the mutants tendril-less and
a®la (apparent late P4 ±P6 and late P2 ±P3 , respectively)
(Meicenheimer et al., 1983; Gould et al., 1986; Gourlay et al.,
2000; Villani and DeMason, 2000), so they may act earlier in
the leaf developmental pathway. No di€erences in the size of
the shoot apical dome have previously been reported in pea
leaf mutants. The smaller diameter (reduced by 25 %) of uni
meristems may be a consequence or a cause of their altered
leaf morphology. It is possible that a reduction in the
number of cells in the incipient leaf primordium of the uni
plant results in the production of a simpler leaf structure.
However, apical size is not always linked to leaf size (Smith
and Hake, 1992; Goliber et al., 1999). The wild-type Uni
gene is not expressed in the shoot apical meristem (Gourlay
et al., 2000), so the reduction in Uni meristem size may not
be a primary e€ect of the mutation. It is possible that cells
expressing Uni in the marginal blastozones normally signal
to cells in the shoot apical meristem. Recent work has
implicated auxin concentration as a determinant of primordial size in tomato, with increasing concentrations of auxin
increasing the number of cells recruited into primordia
(Reinhardt et al., 2000).
CO N C L U S I O N S
The most recent model of pea leaf development (Hofer and
Ellis, 1998) proposes that Uni, with one or more other
genes, acts to maintain a shoot-like fate in the leaf
meristem, with the inhibition of Uni by Af and Tl producing
determinate lea¯ets and tendrils. Hofer and Ellis (1998)
suggest that Coch acts to inhibit Uni at the base of the leaf
to distinguish determinate stipule cells from the partiallyindeterminate rachis. Recent molecular evidence upholds
this model. Gourlay et al. (2000) show that Uni is expressed
in the leaf primordium when leaf primordia are initiated
and is down-regulated when they become determined. The
loss of Uni gene function results in a prematurely
determined leaf primordium which produces a simple leaf
form. In keeping with this, there is prolonged Uni
expression in the primordia of the complex leaf forms
seen in the a®la and a®la tendril-less mutants.
This model is largely supported by the evidence from
mutant phenotypes reported here. A loss of determinacy is
seen in the formation of compound stipules rather than
simple ones in coch plants, and some loss of determinacy
occurs in ¯owers resulting in the production of supernumerary organs. However, the primary e€ect of the coch
mutation seems to be to retard the development of the
stipules; this then appears to allow expression of Uni in
some stipule primordia, leading to their development into
compound structures. The retarded development of the
stipules seems to allow them to behave more like leaf blades.
This is shown by the way coch mutant stipules mimic the
form of the leaf blade (Marx, 1987), and is supported by the
epistasis shown here between coch and st.
We also add information to this model by showing that
uni mutants have a reduced shoot apical meristem diameter,
which could result in their simpler leaf form. However, Uni
234
Yaxley et al.ÐLeaf and Flower Development in Pea
is not expressed in the shoot apical meristem (Gourlay et al.,
2000), so it may act indirectly to maintain shoot meristem
size. This paper also shows reduced action of Tl in the basal
compartment of the leaf, which supports the existence of a
gradient of Tl gene action from high in the distal
compartment to low in the basal compartment, as proposed
by Lu et al. (1996).
AC K N OW L E D G E M E N T S
We thank Jenny Smith, Julian Yaxley, Ian Cummings,
Tracey Jackson, Leanne Sherrif, Fred Koolhof, David
Steele, Rob Wiltshire, Scott Taylor, Shona Batge, Ian
Murfet, Jim Weller, Julie Hofer and Mike Ambrose for
assistance. This work was funded by an ARC grant to JBR
and a GRDC Junior Research Fellowship to JLY.
L I T E R AT U R E C I T E D
Blixt S. 1972. Mutation genetics in Pisum. Agri Hortique Genetica 30:
1±293.
Cote R, Gerrath JM, Posluszny U, Grodzinski B. 1992. Comparative
leaf development of conventional and semilea¯ess peas (Pisum
sativum). Canadian Journal of Botany 70: 571±580.
Crang RFE, Klomparens KL. 1988. Artefacts in biological electron
microscopy. New York: Plenum Press.
Danilatos GD. 1993. Introduction to the ESEM instrument. Microscopy Research and Technique 25: 354±361.
Eriksson G. 1929. Erbkomplexe des Rotklees und Erbsen. Zeitschrift
fuÈr P¯anzenzuchtung 14: 445±475.
Ferrandiz C, Navarro C, Gomez MD, Canas LA, Beltran JP. 1999.
Flower development in Pisum sativum: from the war of the whorls
to the battle of the common primordium. Developmental Genetics
25: 280±290.
Goldenberg JB. 1965. a®la, a new mutation in pea (Pisum sativum L.).
Boletin Genetico 1: 27±31.
Goliber T, Kessler S, Chen J-J, Bharathan G, Sinha N. 1999. Genetic,
molecular, and morphological analysis of compound leaf development. Current Topics in Developmental Biology 43: 259±290.
Gould KS, Cutter EG, Young JPW. 1986. Morphogenesis of the
compound leaf in three genotypes of the pea, Pisum sativum.
Canadian Journal of Botany 64: 1268±1276.
Gould KS, Cutter EG, Young JPW. 1994. The determination of pea
leaves, lea¯ets and tendrils. American Journal of Botany 81:
352±360.
Gould KS, Young JPW, Cutter EG. 1992. L-system analysis of
compound leaf development in Pisum sativum. Annals of Botany
70: 189±196.
Gould KS, Cutter EG, Young JPW, Charlton WA. 1987. Positional
di€erences in size, morphology, and in vitro performance of pea
axillary buds. Canadian Jounal of Botany 65: 406±411.
Gourlay CW, Hofer JMI, Ellis THN. 2000. Pea compound leaf
architecture is regulated by interactions among the genes
UNIFOLIATA, COCHLEATA, AFILA and TENDRIL-LESS.
The Plant Cell 12: 1279±1294.
Hofer JMI, Ellis THN. 1996. The e€ect of Uni on leaf shape. Pisum
Genetics 28: 21±23.
Hofer JMI, Ellis THN. 1998. The genetic control of patterning in pea
leaves. Trends in Plant Science 3: 439±444.
Hofer J, Turner L, Hellens R, Ambrose M, Matthews P, Michael A,
Ellis N. 1997. UNIFOLIATA regulates leaf and ¯ower morphogenesis in pea. Current Biology 7: 581±587.
Hole CC, Hardwick RC. 1976. Development and control of ¯owers per
node in Pisum sativum L. Annals of Botany 40: 707±722.
Jablonski W. 1997. Electron microscopy and X-ray microanalysis: key
research tools in applied science. MSc Thesis, University of
Tasmania, Australia.
Jackson D, Hake S. 1999. Control of phyllotaxy in maize by the
abphyl1 gene. Development 126: 315±323.
Kujala V. 1953. Felderbse bei welcher die ganze Blattspreite in Ranken
umgewandelt ist. Archivum Societatis Zoologicae Botanicae
Fennicae `Vanamo' 8: 44±45.
Lacroix CR, Sattler R. 1994. Expression of shoot features in early leaf
development of Murrya paniculata (Rutaceae). Canadian Journal
of Botany 72: 678±687.
Lamoreaux RJ, Chaney WB, Brown KM. 1978. The plastochron index:
A review after two decades of use. American Journal of Botany 65:
586±593.
Lamprecht H. 1933. Ein unifoliataÐTypus von Pisum mit gleichzeitiger Pistilloidie. Heriditas 18: 56±64.
Lijsebettens MV, Clarke J. 1998. Leaf development in Arabidopsis.
Plant Physiology and Biochemistry 36: 47±60.
Lu B, Villani PJ, Watson JC, DeMason DA, Cooke TJ. 1996. The
control of pinna morphology in wild-type and mutant leaves of
the garden pea (Pisum sativum L.). International Journal of Plant
Science 157: 659±673.
Marx GA. 1987. A suite of developmental mutants that modify pattern
formation in pea leaves. Plant Molecular Biology Reporter 5:
311±335.
Meicenheimer RD, Muelbauer FJ, Hindman FJ, Gritton ET. 1983.
Meristem characteristics of genetically modi®ed pea (Pisum
sativum) leaf primordium. Canadian Journal of Botany 61:
3430±3437.
Molhova E, Vassileva M, Mihailova M. 1988. Cytoembryological study
of cochleata mutant forms with feminized stamens in Pisum
sativum. Genetics and Breeding 21: 377±383.
Pellew C, Sverdrup A. 1923. New observations on the genetics of peas
(Pisum sativum). Journal of Genetics 13: 125±131.
Reinhardt D, Mandel T, Kuhlemeier C. 2000. Auxin regulates the
initiation and radial position of plant lateral organs. The Plant
Cell 12: 507±518.
Sattler R, Rutishauser R. 1992. Partial homology of pinnate leaves and
shoots. Orientation of lea¯et inception. Botanische JahrbuÈcher fuÈr
Systematik P¯anzengeschichte und P¯anzengeographie 114: 61±79.
Smith LG, Hake S. 1992. The initiation and determination of leaves.
The Plant Cell 4: 1017±1027.
Sundberg MD, Orr AR. 1996. Early in¯orescence and ¯oral development in Zea mays landrace Chapalote (Poaceae). American
Journal of Botany 83: 1255±1265.
Sylvester AW, Smith L, Freeling M. 1996. Acquistion of identity in the
developing leaf. Annual Review of Cell and Developmental Biology
12: 257±304.
Tucker SC. 1989. Overlapping organ initiation and common primordia
in ¯owers of Pisum sativum (Leguminosae: Papilionoideae).
American Journal of Botany 76: 714±729.
Villani PJ, DeMason DA. 1997. Roles of AF and TL genes in pea leaf
morphogenesis: characterization of the double mutant
(AFAFTLTL). American Journal of Botany 84: 1323±1336.
Villani PJ, Demason DA. 1999a. The Af gene regulates timing and
direction of major developmental events during leaf morphogenesis in garden pea (Pisum sativum). Annals of Botany 83: 117±128.
Villani PJ, DeMason DA. 1999b. Roles of Af and Tl genes in pea leaf
morphogenesis: Leaf morphology and pinna anatomy of the
heterozygotes. Canadian Journal of Botany 77: 611±622.
Villani PJ, DeMason DA. 2000. Roles of the Af and Tl genes in pea leaf
morphogenesis: shoot ontogeny and leaf development in the
heterozygotes. Annals of Botany 85: 123±135.
Vilmorin PD, Bateson W. 1911. A case of gametic coupling in Pisum.
Proceedings of the Royal Society B 84: 9±11.
Wellensiek SJ. 1959. Neutronic mutations in peas. Euphytica 8:
209±215.
White O. 1917. Studies of inheritance in Pisum. II. The present state of
knowledge of heredity and variation in peas. Proceedings of the
American Philosphical Society 56: 487±588.
Wiltshire RJE, Murfet IC, Reid JB. 1994. The genetic control of
heterochrony: Evidence from developmental mutants of Pisum
sativum L. Journal of Evolutionary Biology 7: 447±465.
Young JPW. 1983. Pea leaf morphogenesis: A simple model. Annals of
Botany 52: 311±316.