Current Biology, Vol. 13, 432–436, March 4, 2003, 2003 Elsevier Science Ltd. All rights reserved. PII S0960-9822(03)00093-9
Female Control of Male Gamete Delivery
during Fertilization in Arabidopsis thaliana
half of the embryo sacs remain unfertilized (Figure 1B),
whereas the remainder develop into normal seeds (Fig-
ure 1C). Remarkably, the micropylar ends of unfertilized
Nicolas Rotman,
1
Fre´de´rique Rozier,
1
Leonor Boavida,
2
Christian Dumas,
1
Fre´de´ric Berger,
1
and Jean-Emmanuel Faure
1,
*
1
Laboratory of Plant Reproduction and embryo sacs contain tangled structures (Figure 1B) not
observed in wild-type plants (Figure 1A).Development
Unite´ Mixte de Recherche 5667 We tested the genetic control of the sire`nemutation to
understand if the diploid tissues of the pistil (sporophyticCentre National de la Recherche Scientifique
Institut National de la Recherche Agronomique generation) or the haploid pollen and/or embryo sac
(gametophytic generation) are defective. We performedUniversite´ Claude Bernard, Lyon I
Ecole Normale Supe´rieure de Lyon reciprocal crosses between heterozygous srn/SRN and
wild-type plants (Table 1). All crosses with srn/SRN pis-Lyon F-69364 Cedex 07
France tils bear embryo sacs with the sire`ne phenotype, in pro-
portions not significantly different from 50% (Table 1).
2
Instituto Gulbenkian de Cieˆncia
PT-2780-156 Oeiras By contrast, pollen from srn/SRN plants crossed with
wild-type pistils leads to normal development (Table 1).Portugal
These results suggest female gametophytic control for
srn. We confirmed this embryo sac control by studying
srn transmission to the next generation (Table 1; chi-
Summary
square test: P
0.2); srn is fully transmitted by the
pollen but rarely by the embryo sac (Table 1). These
Fertilization in both animals and plants relies on the
results exclude the possibility of a sporophytic mutation
correct targeting of the male gametes to the female
acting in the pistils or the ovules and affecting the em-
gametes [1, 2]. In flowering plants, the pollen tube
bryo sacs in a semi-penetrant way. In such a case, srn
carries twomale gametes through thematernal repro-
would indeed be transmitted to half of the progeny when
ductive tissues to the embryo sac, which contains two
wild-type pollen are crossed on srn/SRN pistils, which
female gametes. The pollen tube then releases its two
we didn’t observe (Table 1). Altogether, our data clearly
male gametes into a specialized receptor cell of the
demonstrate that sire`ne is a strict embryo sac mutation.
embryo sac, the synergid cell [2–5]. The mechanisms
The few cases of srn transmission (5.7%) in crosses
controlling this critical step of gamete delivery are
between wild-type pollen and mutant pistils (Table 1)
unknown. Here, data based on the new sire`ne (srn)
could correspond to the rare cases of ovules with a
mutant of Arabidopsis thaliana provide the first evi-
tangled structure and an early seed development (3.5%,
dence for female control over male gamete delivery.
n 579). The sire`ne mutation, corresponding to a single
Live imaging of fertilization shows that wild-type pol-
locus, was mapped between the markers T16K5-TGF
len tubes do not stop their growth and do not deliver
and CDC2BG on chromosome 3.
their contents in srn embryo sacs.
In srn/SRN plants we observed neither a delay in em-
bryo sac development nor major structural anomalies.
Results and Discussion
Under Normarski microscopy, 80.0% (n  260) of the
embryo sacs not yet penetrated by pollen tubes did
Unlike animal sperm cells, the male gametes of flowering
indeed have a mature structure in srn/SRN pistils at
plants are not motile and are, instead, carried by pollen
anthesis, as in wild-type pistils (78.0%, n  387; chi-
tubes [2–5]. As evidenced by genetic data from A. thali-
square test: P
0.5; Figure 1A; [12]). We obtained similar
ana, the embryo sacs attract pollen tubes to bring about
results when observing embryo sacs with confocal laser
fertilization [6–8]. Cell ablation experiments demonstrate
scanning microscopy (CLSM) in pistils fixed 3–5 hr after
that the synergid cells (Figure 1A) are necessary for this
pollination, prior to pollen tube penetration (n  180;
attraction in Torenia fournieri [9]. The pollen tube most
Figure 2A; [13, 14]). However, at 7–9 hr after pollination,
commonly penetrates the embryo sac through one of
srn/SRN pistils showed abnormalities in several embryo
the two synergid cells [10] and discharges its contents,
sacs that had just been penetrated by a pollen tube (n
including the two sperm cells, next to the two female
23; Figures 2B and 2C). Unlike the situation in the wild-
gametes, the egg cell and the central cell [3–5, 11]. To
type [4, 15], the pollen tube grows deeply into the em-
date, the molecular and cellular mechanisms controlling
bryo sac, between the egg and synergid cells; it almost
pollen tube penetration into the embryo sac and subse-
reaches the central cell and sometimes forms a loop.
quent male gamete delivery are unknown [4]. In order
Also, neither of the two synergid cells shows visible
to isolate mutants defective in these steps of fertiliza-
alteration of its structure (Figures 2B and 2C), whereas
tion, we performed a cytological screen for mutants in
one degenerates in wild-type embryo sacs, either shortly
which seeds fail to develop in the absence of structural
before pollen tube penetration or after pollen tube dis-
anomalies in the embryo sacs and pollen. In pistils of
charge (Figure 2D; [13, 14]). These observations indicate
the sire`ne (srn) mutant at 2 days after pollination (DAP),
that early interaction between the pollen tube and the
embryo sac is altered in sire`ne mutants. We further ob-
served overgrowth of the pollen tube, stained with ani-*Correspondence: jean-emmanuel.faure@ens-lyon.fr

Brief Communication
433
embryo sacs may also be attractive for supernumerary
pollen tubes.
To further understand how the interactions between
the pollen tube and the embryo sac are disrupted by
srn, we developed an original imaging system in A. thali-
ana. For the first time in plants, observations of pollen
tube growth and discharge were made within a pistil.
Microdissected pistils pollinated in planta with EGFP-
expressing pollen [14, 16] are mounted in microcham-
bers (Figures 4A–4C) and imaged with CLSM. In the
wild-type, one pollen tube penetrates the micropyle of
an ovule, reaches the embryo sac (Figure 4D), and forms
an extension in it (Figure 4D, 4 min panel). The pollen
tube releases its contents into the embryo sac (Figure
4E; Movie 1, available with this article online) within less
than a minute (n  26 out of 29), and probably less
than 20 s (based on three recordings with a higher time
resolution). The explosive discharge observed in the
pistils of A. thaliana is consistent with previous observa-
tions of in vitro ovule/pollen tube interaction in T. four-
nieri [11]. In srn/SRN pistils, however, some wild-type
Figure 1. Embryo Sac Structure, as Observed with Normarski Mi-
pollen tubes do not stop their growth once they enter
croscopy after Clearing of A. thaliana Ovules
the embryo sac (n 15; Figure 4E; Movies 2–4, available
(A) Unfertilized embryo sac in a wild-type pistil at anthesis. It con-
with this article online). They coil up in the micropylar
tains a central cell with a single nucleus (nucleolus: ccn), and a
highly polarized egg cell (nucleolus: ecn). One of the two synergid part and produce the tangled structure observed in fixed
cells is visible (nucleolus: syn), with a characteristic polarity opposite
srn/SRN pistils. Also, they do not release their contents
to that of the egg cell. “ot” indicates the outer integuments of the
during up to 6.5 hr of time-lapse imaging. We conclude
ovule.
that srn embryo sacs are unable to stop the growth of
(B) Unfertilized embryo sac in a srn/SRN pistil 2 days after pollina-
pollen tubes and induce their discharge. Interestingly,
tion. Tangled material (tm) is visible in the micropylar part. There is
pollen tube overgrowth is restricted to the micropylar
no sign of fertilization; we can see the egg cell nucleus (nucleolus:
ecn) as well as the central cell nucleus (nucleolus: ccn). part of the srn embryo sac. This may be due either to
(C) Wild-type-looking ovules in a srn/SRN pistil at 2 days after polli-
structural constraints or to the persistence of a synergid
nation are fertilized: The embryo proper is at a two-cell stage (emb),
cell signal attracting the pollen tube tip [9]. Time-lapse
and the endosperm (en) developed into a syncytial structure con-
recordings also confirmed the occurrence of multiple
taining several nuclei (nucleoli: arrow-heads). The scale bar repre-
pollen tube entries in ovules bearing sire`ne embryo sacs
sents 10 m.
(n  3; Movie 4, available with this article online). We
rarely observed such cases of multiple penetrations
line blue, in srn/SRN pistils collected 1 day after anthesis when imaging wild-type ovules. In 94.7% of the ovules
(Figures 3A–3C). We hypothesized that sire`ne embryo with two or three pollen tubes on the funiculus (n 
sacs are penetrated by pollen tubes that coil and maybe 19), only one pollen tube was seen to penetrate the
branch into the micropylar part (Figure 3D). In 17.5% micropyle and to discharge its contents into the embryo
(n  471) of the ovules bearing embryo sacs with coiled sac. After the initial pollen tube had entered the micro-
tubes, we could detect two or more pollen tubes (Figure pyle, supernumerary pollen tubes were seen to stop
3C). This is significantly different from the wild-type situ- growing, change growth direction, or branch (Movie 5,
ation, in which only 5.8% of the embryo sacs are pene- available with this article online; also see Movie 1).
trated by multiple pollen tubes (n 611; chi-square test: Altogether, our data clearly demonstrate that the fe-
male embryo sac controls the step of male gamete deliv-P  10
10
). Thus, unlike wild-type embryo sacs, sire`ne
Table 1. Segregation of the srn Mutation
srn/SRN
a
 SRN/SRN SRN/SRN
a
 srn/SRN srn/SRN
a
 srn/SRN
Wild-type seeds
b
51.7% ( n  453
c
) 100% (n  549
c
) 48.8% (n  705
c
)
d
Wild-type plants in F1
e
94.3% (n  246
f
) 41.1% (n  438
f
) 34.4% (n  163
f
)
g
a
The first plant indicated in the cross is the female partner.
b
We scored wild-type seeds 2 days after hand pollination by clearing the pistils and observing the structures under Normarski microscopy.
c
Total number of wild-type seeds and srn ovules observed.
d
The proportions of wild-type seeds are not significantly different from the theoretical proportions for a female gametophytic mutant (P

0.85, chi-square test).
e
Seeds from each cross were collected and germinated. These F1 plants were then screened for semi-sterility. The srn phenotype of the
plants with semi-sterility could be confirmed by Normarski microscopy.
f
Total number of plants screened.
g
The proportions of wild-type plants are not significantly different from the theoretical proportions for a female gametophytic mutant with a
low (5.7%) female transmission of the mutation (P
0.2, chi-square test).