The 14-3-3 Protein GF14c Acts as a Negative Regulator of

Flowering in Rice by Interacting with the Florigen Hd3a

Yekti Asih Purwestri, Yuka Ogaki, Shojiro Tamaki, Hiroyuki Tsuji and Ko Shimamoto*

Rapid Paper
Laboratory of Plant Molecular Genetics, Nara Institute of Science and Technology, 8916-5 Takayama, Ikoma, Nara, 630-0101 Japan

Hd3a and FT proteins have recently been proposed to act Introduction

as florigens in rice and Arabidopsis, respectively; however,
the molecular mechanisms of their function remain to be Flowering is critical for growth and reproduction in plants,

Downloaded from [Link] at University of Windsor on October 19, 2014

determined. In this study, we identified GF14c (a 14-3-3 and is controlled by both environmental and endogenous
protein) as an Hd3a-interacting protein in a yeast two- conditions. One of the most important factors that controls
hybrid screen. In vitro and in vivo experiments, using a flowering is the plant's response to photoperiod (Imaizumi
combination of pull-down assays and bimolecular and Kay 2006). In addition to the photoperiodic pathway,
fluorescence complementation, confirmed the interaction the regulation of flowering time involves complex signaling
between Hd3a and GF14c. Functional analysis using either pathways, such as the vernalization, autonomous flowering
GF14c overexpression or knockout transgenic rice plants and gibberellins pathways. Arabidopsis thaliana and Oryza
indicated that this interaction plays a role in the regulation sativa are used as models to study the regulation of flower-
of flowering. GF14c-overexpressing plants exhibited a ing time in long-day (LD) plants and short-day (SD) plants,
respectively. Three genes that constitute the major genetic
delay in flowering and the knockout mutants displayed
pathway in the photoperiodic regulation of flowering in rice
early flowering relative to the wild-type plants under
have been isolated: OsGI (O. sativa GIGANTEA), an ortholog
short-day conditions. These results suggest that GF14c
of Arabidopsis GI; Hd1 (Heading date 1), an ortholog of Ara-
acts as a negative regulator of flowering by interacting
bidopsis CO (CONSTANS); and Hd3a (Heading date 3a), an
with Hd3a. Since the 14-3-3 protein has been shown to
ortholog of Arabidopsis FT (FLOWERING LOCUS T). The
interact with FT protein in tomato and Arabidopsis, our major difference between rice, an SD plant, and Arabidopsis,
results in rice provide important findings about FT an LD plant, is the regulation of Hd3a/FT by Hd1/CO. Under
signaling in plants. LD conditions, this regulation is positive in Arabidopsis but
Keywords: Flowering • GF14c • Hd3a • Protein interaction • negative in rice (Hayama et al. 2003, Izawa 2007, Tsuji et al.
Rice. 2008).
Hd3a was first identified as a quantitative trait locus
Abbreviations: 3-AT, 3-aminotriazole; BiFC, bimolecular (QTL) that promotes flowering of rice under SD conditions
fluorescence complementation; bZIP, basic/leucine zipper; (Yamamoto et al. 1998, Monna et al. 2002). Overexpression
CaMV, cauliflower mosaic virus; DTT, dithiothreitol; GFP, of Hd3a protein with a constitutive promoter (Kojima et al.
green fluorescent protein; GST, glutathione S-transferase; 2002) or vascular-specific promoters (Tamaki et al. 2007)
GUS, β-glucuronidase; LD, long day; PEBP, phosphatidyl- results in an early-flowering phenotype, and suppression of
ethanolamine-binding protein; QTL, quantitative trait locus; Hd3a with RNA interference (RNAi) delays flowering Editor-in-Chief’s choice
RKIP, Raf kinase inhibitor protein; RNAi, RNA interference; (Komiya et al. 2008). Hd3a is a member of a large gene family
RT–PCR, reverse transcription–PCR; SAM, shoot apical consisting of at least 13 genes in the rice genome (Chardon
meristem; SD, short day; TBS, Tris-buffered saline; ZT, and Damerval 2005), and at least two paralogs, RFT1 (Rice
zeitgeber time. FLOWERING LOCUS T1) and FTL (FT-Like), are reported to

*Corresponding author: E-mail, simamoto@[Link]; Fax, +81-743-72-5502.

Plant Cell Physiol. 50(3): 429–438 (2009) doi:10.1093/pcp/pcp012, available FREE online at [Link]
© The Author 2009. Published by Oxford University Press on behalf of Japanese Society of Plant Physiologists.
All rights reserved. For permissions, please email: [Link]@[Link]

Plant Cell Physiol. 50(3): 429–438 (2009) doi:10.1093/pcp/pcp012 © The Author 2009. 429
 Y. A. Purwestri et al.

have a flowering promotion function (Izawa et al. 2002, Indeed, a mouse PEBP, RKIP, has been shown to interact with
Komiya et al. 2008). and inhibit the activity of the kinase Raf1 (Yeung et al.
Hd3a expression is regulated by Hd1 and Ehd1 (Early 1999).
heading date1) that encodes a B-type response regulator and Most recently, Hd3a and its ortholog have been proposed
is a unique flowering time gene in rice (Doi et al. 2004). Ehd1 as florigens, or mobile flowering signals (Corbesier et al. 2007,
promotes floral transition preferentially under SD condi- Jaeger and Wigge et al. 2007, Lin et al. 2007, Mathieu et al.
tions, even in the absence of functional alleles of Hd1. Expres- 2007, Tamaki et al. 2007, Notaguchi et al. 2008, Tsuji et al.
sion analysis revealed that Ehd1 functions upstream of Hd3a, 2008). The next question that needs to be addressed is the
RFT1 and some MADS-box genes (Doi et al. 2004). More mechanism of Hd3a function. One important step in the
recently, Ghd7, which encodes a CCT- (CO, CO-LIKE and characterization of Hd3a function is to identify other pro-
TIMING OF CAB1) domain protein, was isolated based on teins with which it interacts.
natural variation in rice (Xue et al. 2008). Ghd7 affects levels A study in tomato revealed several SP-interacting pro-
of Ehd1 and Hd3a transcripts, but does not affect Hd1 mRNA teins including a 14-3-3 family member, a protein kinase and

Downloaded from [Link] at University of Windsor on October 19, 2014

levels. Ghd7 represses Ehd1 and Hd3a expression under LD a bZIP transcription factor (Pnueli et al. 2001). SFT also inter-
conditions, leading to delayed flowering. Therefore, two acts with 14-3-3, as well as with bZIP (Pnueli et al. 2001).
independent floral pathways are present in rice: the con- In Arabidopsis, FT interacts with FD and 14-3-3 proteins
served Hd1 pathway and a unique Ehd1 pathway that may (Abe et al. 2005, Pnueli et al. 2001, Wigge et al. 2005); how-
integrate environmental photoperiod signals into the ever, no Hd3a-interacting proteins have yet been identified
expression of FT-like genes (Izawa 2007, Tsuji et al. 2008). in rice.
Recently, another flowering gene, RID1/OsId1/Ehd2, was In this study, we identified GF14c (a G-box factor 14-3-3c
identified in rice. This gene encodes a putative transcription protein) as an Hd3a-interacting protein both in vitro and
factor with a zinc finger motif orthologous to ID1 (INDETER- in vivo. Since 14-3-3 family members interact with diverse
MINATE1) that promotes flowering in maize (Zea mays) proteins, it is of interest to understand their role in Hd3a
(Matsubara et al. 2008, Park et al. 2008, Wu et al. 2008). signaling. Functional analyses using both knockout and over-
RID1/OsId1/Ehd2 promotes the floral transition mainly by expressing plants indicate that GF14c acts as a negative
up-regulating Ehd1 and genes downstream of Ehd1, such as regulator of flowering in rice.
Hd3a and RFT1.
FT encodes an approximately 23 kDa protein whose
sequence is similar to that of Raf kinase inhibitor protein Results
(RKIP) and phosphatidylethanolamine-binding protein Hd3a interacts with GF14c in a yeast two-hybrid
(PEBP) (Ahn et al. 2006). The PEBP family regulates signaling
system
pathways to control growth and differentiation. PEBPs seem
to act biochemically as inhibitors, binding signaling compo- To understand the molecular function of Hd3a, we perfor-
nents to reduce the flux through their pathways (Hanzawa med a yeast two-hybrid screen to search for Hd3a-interacting
et al. 2005). The crystal structures of PEBP from human and proteins, using full-length Hd3a as bait for a cDNA library
bovine sources (Banfield et al. 1998) and CENTRORADIALIS constructed from leaf blades harvested when Hd3a expres-
from Antirrhinum (Banfield and Brady 2000), as well as those sion was induced. From 1.6×106 yeast transformants, 96
of Arabidopsis [TERMINAL FLOWER1 (TFL1) and FT] (Ahn clones grew on selective medium lacking histidine. Three
et al. 2006), indicate that the ligand-binding pocket is acces- independent clones were identified as GF14c, an isoform of
sible for interacting with protein partners. PEBPs may also the rice 14-3-3 gene family (Table 1). Only a partial fragment
act as either scaffolds for or regulators of signaling com- in the C-terminal region (115–256) was identified originally
plexes, as suggested by the finding that SP (SELF PRUNING) from the yeast-two hybrid screen. An experiment using full-
and SFT (SINGLE-FLOWER TRUSS), the tomato orthologs of length Hd3a and full-length GF14c, either as bait or as prey,
TFL1 and FT, respectively, can interact with diverse proteins further confirmed that Hd3a and GF14c interact in the yeast
(Pnueli et al. 2001, Lifschitz et al. 2006). In Arabidopsis, FT system (Fig. 1). In plants, 14-3-3 proteins have been known
interacts with the basic/leucine zipper (bZIP) transcription to interact with Hd3a orthologs: FT in Arabidopsis and SP
factor FD (Abe et al. 2005, Wigge et al. 2005). The bZIP tran- and SFT in tomato (Pnueli et al. 2001). GF14 family members
scription factor SPGB in tomato, a homolog of FD, also inter- have roles in both disease resistance and development in
acts with SP and SFT. Structural analysis of CEN indicates rice (Cooper et al. 2003). We are interested in further charac-
that its ligand-binding pocket is incapable of accommodat- terizing the molecular function of GF14c in association with
ing phosphoryl groups; however, PEBPs may mediate signal- Hd3a throughout rice development, particularly during the
ing via their association with phosphorylated proteins. floral transition.

430 Plant Cell Physiol. 50(3): 429–438 (2009) doi:10.1093/pcp/pcp012 © The Author 2009.
 GF14c interacts with the florigen Hd3a in rice

Table 1 Isolation of GF14c, an Hd3a-interacting protein, in a yeast two-hybrid screen

Number Accession No. Homologous protein Two hybrida Amino acidsa
IV-B1 gi 50903393 GF14-c protein [Oryza sativa (japonica cultivar group)]
VI-G2 AK122149/Os8g0430500 14-3-3 homologous 141 (115–256) 256
VI-A3
aLength of amino acids.

Bait/Prey +H –H –H + 3-AT abundantly during plant development. We also examined

the distribution of Hd3a and 14-3-3 proteins in rice plants.
empty/empty Total proteins from Hd3a-overexpressing plants under its
native promoter (Hd3a::Hd3a:GFP) (Tamaki et al. 2007) and

Downloaded from [Link] at University of Windsor on October 19, 2014

non-transgenic control plants co-expressed Hd3a and 14-3-3
Hd3a/empty
proteins in both leaf blades and stems (Fig. 2C).

Subcellular localization of Hd3a and GF14c

empty/GF14c
To gain insight into the molecular function of Hd3a and
GF14c, we made a fusion construct to express mCherry fluo-
Hd3a/GF14c rescent protein-linked Hd3a under the ubiquitin promoter
and green fluorescent protein (GFP)-linked GF14c driven by
the caulifower mosaic virus (CaMV) 35S promoter, and ana-
GF14c/empty
lyzed the intracellular localization of Hd3a and GF14c. These
constructs were introduced into rice protoplasts isolated
empty/Hd3a from rice suspension culture cells. All rice protoplasts
observed in this experiment showed that Hd3a–mCherry
was associated with both the cytoplasm and nucleus; how-
GF14c/Hd3a ever GFP–GF14c predominantly localized in the cytoplasm
(75% of rice protoplasts observed) (Fig. 2D). The predomi-
Fig. 1 Identification of GF14c as an Hd3a-interacting protein in yeast.
nant cytoplasmic localization of 14-3-3 proteins may suggest
Growth of yeast colonies on plates lacking histidine (–H) or lacking that they function as cytoplasmic anchors that regulate the
histidine plus 2.5 mM 3-aminotriazole (3-AT) indicates a positive import of proteins into the nucleus or other organelles or
interaction between Hd3a and GF14c. Two independent clones are promote export from the nucleus to the cytoplasm (Muslin
shown for each treatment. and Xing 2000).

In vitro interaction of Hd3a and GF14c

Tissue-specific expression of Hd3a and GF14s in rice
To study the interaction between Hd3a and GF14c, we
From the International Rice Annotation Project Database performed a glutathione S-transferase (GST) pull-down
(Rice Annotation Project, 2008), we identified GF14 family assay. A GST–Hd3a fusion protein was pulled down with
members. GF14s belong to the 14-3-3 protein group and His-tagged GF14c, as shown in Fig. 3A, indicating that Hd3a
have eight isoforms, designated as GF14a–h in rice (Chen interacted with GF14c in vitro. Results of this experiment
et al. 2006a, Yao, et al. 2007), that share 85–95% amino acid were thus consistent with the results of the yeast two-hybrid
identity (Fig. 2A). To determine whether the GF14 genes are experiment.
expressed in rice organs, we performed semi-quantitative
reverse transcription–PCR (RT–PCR) using specific primers The Hd3a and GF14c complex is localized in the
(Fig. 2B). The GF14 genes were expressed in all tested organs, cytoplasm by BiFC
suggesting that they have broad functions in rice growth We used bimolecular fluorescence complementation (BiFC)
and development. The expression levels of GF14 a–f seem to determine the distribution of Hd3a and GF14c in vivo.
to be higher than those of GF14g and GF14h. No diurnal The expression vectors for Hd3a protein fused to the
changes or developmental patterns of GF14c expression N-terminal half of mVenus (Vn) and GF14c protein fused to
were observed (Supplementary Fig. S1), indicating that the C-terminal half of mVenus (Vc) were transiently intro-
GF14c is expressed independently of the photoperiod and duced into the rice protoplasts, with the mCherry expression

Plant Cell Physiol. 50(3): 429–438 (2009) doi:10.1093/pcp/pcp012 © The Author 2009. 431
 Y. A. Purwestri et al.

C FP
A a :G
Os08g0430500 GF14c pe d3
ty :H
- a:
ild 3
Os02g0580300 GF14e W Hd
Os04g0462500 GF14b es es s
av eav em
Os03g0710800 GF14f le l st
Hd3a-GFP
Os11g0546900 GF14d
Os08g0480800 GF14a 14-3-3
Os01g0209200 GF14g
CBB
Os11g0609600 GF14h

B D
s

ex

Downloaded from [Link] at University of Windsor on October 19, 2014

th
sh s

ap
de
gs

ca e a
a
lin

sh s
ts
s

bl

r
we
s
ed

ed

em

oo
ot

lli
af

af
se

se

flo
ro

st

le

le

GF14a
GF14b

GF14c
GF14d
GF14e
GF14f
GF14g
GF14h
Ubq

Fig. 2 Tissue-specific and subcellular distribution of Hd3a and GF14s family members in rice. (A) A phylogenetic tree was constructed with the
ClustalW program using the neighbor-joining method. The eight GF14 isoforms share a high degree of amino acid identity (85–95%). (B) Rice
GF14 family members are expressed in all organs tested. Using gene-specific primers, RT–PCR was performed with 1 µg of total RNA extracted
from various organs of the wild-type N8 cultivar. (C) Immunoblot detection of cytosolic fractions of wild-type and Hd3a-ox plants. Hd3a and
GF14s were detected by a GFP and by a 14-3-3 antibody, respectively. Coomassie brilliant blue staining (CBB) was used as a loading control. (D)
GF14c is localized mainly in cytosol. Transient expression of GFP–GF14c and Hd3a–mCherry in rice protoplasts was driven by the CaMV 35S and
ubiquitin promoters, respectively. NLS-2xmOrange was used as a marker for the nucleus. mCherry was used as a marker of the nucleus and
cytoplasm. Upper panels: co-expression of GFP–GF14c and Hd3a–mCherry proteins. Lower panels: co-expression of GFP–GF14c and mCherry
proteins. Bar = 10 mm.

plasmid as a marker for transformed cells. As shown in Fig. 3B, phenotypic analysis of GF14c. GF14c-ox plants were delayed
95% of transformed rice protoplasts observed in this experi- by 5–20 d in flowering relative to wild-type plants (Fig. 4B,
ment showed Hd3a and GF14c interaction in both protein C). A t-test confirmed that the flowering time was signifi-
fusion combinations as documented by a strong green fluo- cantly different between GF14c-ox and wild-type plants
rescence concentrated in the cytoplasm. Hd3a-Vn/GUS-Vc (P = 0.0046). Other phenotypes observed in this mutant
or GUS-Vn/GF14c-Vc combination showed remarkably few were erect leaves and longer culms compared with the wild-
numbers of cells with Venus fluorescence (0–10% of the type plants (data not shown). In GF14c-ox plants, overex-
transformed cells). These results clearly demonstrate that pression of GF14c at both the mRNA and the protein level
Hd3a interacts with GF14c mainly in the cytoplasm. was confirmed (Fig. 4D, E).
To study further the function of GF14c in flowering, we
GF14c-overexpressing and gf14c mutant plants also used the gf14c/GF14c heterozygous T-DNA insertion
To characterize the function of GF14c, a GF14c overexpres- mutants in japonica rice cv ‘Dongjin’ generated by Postech,
sion (ox) construct driven by the CaMV 35S promoter was Korea. One heterozygous line (Fig. 5B) possessed a T-DNA
generated and transformed into rice plants (Fig. 4A). T1 insertion in the second exon of GF14c (Fig. 5A, D). We ger-
generation transgenic plants (n = 15) were used for the minated 50 seeds derived from the gf14c T-DNA mutants

432 Plant Cell Physiol. 50(3): 429–438 (2009) doi:10.1093/pcp/pcp012 © The Author 2009.
 GF14c interacts with the florigen Hd3a in rice

A B BiFC mCherry Bright Field

input pull down
Hd3a-Vn
GST + − − + − +
GST-Hd3a − + − − + GF14c-Vc
His-GF14c − − + + + Nuc

CBB
GUS-Vn
+
GF14c-Vc

Hd3a-Vn
+
α-His

Downloaded from [Link] at University of Windsor on October 19, 2014

GUS-Vc

Fig. 3 In vitro and in vivo interaction of Hd3a with GF14c. (A) Hd3a–GST interacts with GF14c–His in vitro. GF14c-tagged His was pulled down
by Hd3a-tagged GST. GF14c was detected with an α-His antibody. (B) BiFC analysis of the interaction between Hd3a and GF14c in rice protoplasts.
Bars = 10 µm.

A C 9
8
35S cDNA GF14c nos
7

6
Plant number

5
B
4 WT
panicle 3 GF14c ox
2
1

0
57 58 59 60 61 63 65 69 70 82 83
Days to heading (days after sowing)

GF14c ox
D WT #1 #2

hpt
E GF14c ox
GF14c (trans)
WT #1 #2

GF14c (endo) 14-3-3

Hd3a Tubulin

ubq

WT #1 #10
GF14c -ox

Fig. 4 Characterization of GF14c-overexpressing (GF14c-ox) plants. (A) GF14c expression was controlled by the CaMV 35S promoter. (B) Growth
of N8 wild-type and GF14c-ox plants under SD conditions. The arrow indicates a panicle. (C) Flowering time of GF14c-ox plants (GF14c ox) was
delayed under SD conditions. (D) GF14c and Hd3a mRNA levels in wild-type and GF14c-ox plants. 1 and 2 indicate independent transgenic lines.
The GF14c transgene was detected in transgenic plants. Hd3a levels in transgenic plants were not changed compared with the wild type. (E)
GF14c protein levels in GF14c-ox plants. Tubulin was used as a loading control.

Plant Cell Physiol. 50(3): 429–438 (2009) doi:10.1093/pcp/pcp012 © The Author 2009. 433
 Y. A. Purwestri et al.

A D
P1 P2 P3
gf14c/GF14c
RB LB
TAA WT #1 #2
ATG
P1-P2

P1-P3
B

panicle E
gf14c/GF14c
WT #1 #2
C GF14c

Downloaded from [Link] at University of Windsor on October 19, 2014

80 Hd3a

Days to heading (days after sowing)

70 Ubq

60

50 F gf14c/GF14c

40 WT #1 #2

30 14-3-3

20
CBB
10

0
WT gf14c/ gf14c/
GF14c-1 GF14c-3 WT gf14c/GF14c

Fig. 5 Characterization of gf14c T-DNA insertion plants. (A) Structure of genomic GF14c (Os08g0430500) and position of the T-DNA insertion.
GF14c consists of five exons (filled boxes) and four introns (lines). The T-DNA was inserted into the second exon in the opposite orientation to
the direction of GF14c transcription. Arrows indicate primers used for analyzing the T-DNA insertion site. LB and RB represent the left and right
borders of T-DNA, respectively. (B) Growth of wild-type and two gf14c mutant plants (cv ‘Dongjin’) under SD conditions. Arrows indicate panicles.
(C) Heading date of gf14c mutant and wild-type plants under SD conditions. Error bars represent the SEM. (D) PCR analysis of GF14c with primers
(Supplementary Table S1) indicated in (A). (E) GF14c and Hd3a mRNA levels in wild-type and GF14c-expressing plants. (F) Expression level of
GF14c protein in GF14c-ox plants. Coomassie brilliant blue staining (CBB) was used as a loading control.

and found that 24 were heterozygous, seven were wild type apical meristem (SAM). Since the tissue localization of Hd3a
and 19 did not germinate, which were likely to be homozy- protein was determined using transgenic rice plants express-
gous, suggesting that the mutation may be lethal. The gf14c/ ing Hd3a–GFP fusion protein driven by the Hd3a promoter,
GF14c plants flowered earlier than those of the wild type by and Hd3a protein was present in vascular tissue from the leaf
about 14 d (12–17-d depending on the plants) (Fig. 5C). A blades and transported through the phloem (Tamaki et al.
t-test confirmed that the flowering time was significantly 2007), we prepared a cDNA library from leaf blades to search
different between the wild type and gf14c (P = 0.0387, n = 5). for Hd3a-interacting proteins in a yeast two-hybrid screen.
GF14c expression at both the mRNA and protein levels in We identified GF14c, a member of the 14-3-3 protein family
these mutants plants was lower than in wild-type plants whose members are highly conserved in all eukaryotes (Ferl
(Fig. 5E, F). These results clearly indicate that GF14c is 1996). The 14-3-3 proteins are well known as proteins that
involved in regulation of flowering time in rice. are involved in signaling pathways in eukaryotes. In plants,
14-3-3 proteins have been shown to interact with FT and
TFL1 in Arabidopsis, and with SP and SFT in tomato (Pnueli
Discussion
et al. 2001, Lifschitz et al. 2006). More recently, a comprehen-
Hd3a is thought to move along leaf cells and the vascular sive study using yeast two-hybrid and affinity chromatography
system of leaves and stems, and exerts its action in the shoot with mass spectrometry provided a large number of novel

434 Plant Cell Physiol. 50(3): 429–438 (2009) doi:10.1093/pcp/pcp012 © The Author 2009.
 GF14c interacts with the florigen Hd3a in rice

Cells of the SAM

Hd3a
flowering

Hd3a
Hd3a
OsFD?
OsMADS14/15
nucleus

Hd3a

Downloaded from [Link] at University of Windsor on October 19, 2014

Hd3a

Hd3a
GF14c

Hd3a Hd3a
Companion cells

Fig. 6 Possible role of GF14c in Hd3a signaling in rice flowering. As the floral stimulus that controls floral transition in the shoot apical meristem
(SAM), Hd3a has the capacity to traffic cell to cell and move long distances via the phloem. Hd3a is a small protein with molecular weight 23 kDa,
thus it may be able to move from companion cells to sieve elements. The overexpression of GF14c may cause inhibition of Hd3a trafficking to
SAM and leads to a delay in flowering. In the absence of GF14c, Hd3a may move freely into the nucleus and activates floral induction.

putative 14-3-3-interacting partners in barley (Schoonheim only a very weak fluorescent signal was observed in the
et al. 2007). Interestingly, a wide range of 14-3-3 targets have nucleus (Fig. 2D). To address the question of whether the
functions in various signal transduction pathways and might interaction between Hd3a and GF14c would inhibit shut-
form a platform for cross-talk in the signal transduction tling of Hd3a from the cytoplasm into the nucleus, we per-
pathway. Binding of 14-3-3 proteins regulates their partner formed a BiFC experiment. The results demonstrate that the
proteins through a variety of mechanisms, such as altering Hd3a–GF14c protein complex is mainly localized in the
their catalytic activity, subcellular localization, stability or cytoplasm (Fig. 3B). This result strongly suggests that Hd3a
their interaction with other proteins (Roberts, 2003). and GF14c interaction increases Hd3a cytoplasmic reten-
Identification of GF14c as an Hd3a partner has been tion. Therefore, the delayed flowering phenotype in GF14c-ox
confirmed by several methods. The results from yeast two- plants can be explained by the increased cytoplasmic reten-
hybrid assays using a full-length construct, in vitro pull-down tion of Hd3a by GF14c (Fig. 6). In Arabidopsis, to initiate
and BiFC assays demonstrated the interaction between floral transition in the shoot apex, FT interacts with FD, a
Hd3a and GF14c in rice. The subcellular distribution of bZIP transcription factor that localizes in the nucleus to
GF14c demonstrated their localization in both the cyto- induce target meristem identity genes such as APETALA1
plasm and nucleus. The 14-3-3 proteins are highly conserved (AP1) (Abe et al. 2005). This process could be attenuated by
throughout the animal and plant kingdom. In Arabidopsis, the cytoplasmic retention of Hd3a by GF14c. Several lines of
there are at least 12 isoforms, whereas eight isoforms exist in evidence indicate the function of 14-3-3 protein in nuclear–
rice. These proteins are known as a component associated cytoplasmic shuttling in the signal transduction pathway
with a number of different proteins in signal transduction (Igarashi et al. 2001, Ishida et al. 2004).
pathways, including plant hormone signaling pathways. Another mechanism which could possibly explain the
Overexpression of GF14c caused late flowering, suggest- phenotypes of GF14c-ox plants and the gf14c knockout
ing that 14-3-3 functions as a negative regulator of flowering. mutants is the interaction of Hd3a with GF14c, which inhib-
Hd3a protein is localized in both the cytoplasm and nucleus; its movement of Hd3a in the long-distance trafficking
however, GF14c is mainly localized in the cytoplasm and through the phloem (Fig. 6). Interestingly, the gf14c knockout

Plant Cell Physiol. 50(3): 429–438 (2009) doi:10.1093/pcp/pcp012 © The Author 2009. 435
 Y. A. Purwestri et al.

mutants exhibited an early flowering phenotype. These with Hd3a. A swapping experiment, using full-length
mutants also showed a dwarf phenotype and increased tiller constructs of GF14c and Hd3a, either as bait or as prey, was
numbers (Supplementary Fig. S2), suggesting that GF14c performed. Interactions were tested on SC medium lacking
could also function independently of flower induction at the histidine (–H) or lacking histidine and containing 2.5 mM
SAM. When expression levels of GF14c are low in the leaf, 3-aminotriazole (3-AT).
Hd3a protein may be capable of moving long distances freely
from companion cells to the phloem sieve elements since Full-length cDNA construction
Hd3a is a small protein that is under the size exclusion limit Since only a partial cDNA was obtained from yeast two-
of plasmodesmata (Lough and Lucas 2006, Giakountis hybrid screening, we constructed a full-length GF14c cDNA
and Coupland 2008). In contrast to our results in rice, in based on sequence information in the rice DNA database.
Arabidopsis, loss of 14-3-3 µ and 14-3-3 υ functions pro- We designed forward and reverse primers (see Supplementary
duced a slight delay in flowering (3–5 d) in comparison with Table S1) to produce a full-length PCR fragment using cDNA
the wild type (Mayfield et al. 2007). These authorts also as a template. The PCR fragment was then subcloned into

Downloaded from [Link] at University of Windsor on October 19, 2014

tested the interaction with components of the photoperi- the pENTR/D-TOPO cloning vector (Invitrogen, Carlsbad,
odic pathway, such as CO, GI, ZTL, PIF3 and TOC1; however, CA, USA) to obtain an entry clone.
the interaction with FT was not tested. This previous study
also showed that 14-3-3 µ and 14-3-3 υ interacted with CO, Database analysis of GF14c
a central regulator of the photoperiod pathway (Mayfield
Rice GF14 family sequences were obtained from the
et al. 2007). Therefore, although our study demonstrates an
GenBank database and the International Rice Annotation
interaction between Hd3a and GF14c, and indicates that
Project Database (Rice Annotation Project, 2008). For data
GF14c acts as a negative regulator of flowering in rice, molec-
mining of GF14c, the deduced amino acid sequences were
ular mechanisms of GF14c function in floral transition and
aligned using ClustalW with default parameters, and a phy-
development remain to be studied.
logenetic tree was generated using the neighbor-joining
method.
Materials and Methods
Plant materials and growth conditions In vitro pull-down assay
Japonica rice cultivars ‘Norin 8’ (N8) and ‘Dongjin’ were used Hd3a and GF14c were cloned into pDEST15 and pDEST17
as wild-type plants. GF14c-overexpressing plants (35S::GF14c) vectors (Invitrogen), respectively. GST, GST–Hd3a and
were generated by Agrobacterium-mediated transformation His–GF14c were expressed in Escherichia coli BL21 (DE3)
in an N8 background. T-DNA insertion alleles of GF14c in a (Invitrogen) and purified with glutathione–agarose (Sigma,
‘Dongjin’ background (IB-10131) were obtained from Tokyo, Japan) and HisTrap (Amersham) columns according
Dr. Gynheung An, POSTECH, Korea. To confirm insertion to the manufacturers’ instructions. In vitro binding assays
sites, the sequences flanking the right border of T-DNA were were performed as follows. The concentration of each fusion
determined using primers designed from sequence informa- protein was determined by Coomassie staining. Equal
tion provided in the POSTECH database. To determine flow- amounts of GST–Hd3a protein coupled to glutathione–
ering time phenotypes, all plants were grown in climate Sepharose 4B beads and His–GF14c were incubated in TEDM
chambers under SD conditions with a 24 h temperature buffer [10 mM Tris–HCl, pH 7.5/0.5 mM EDTA, pH 7.5/1 mM
cycle (10 h, 30°C during subjective day; 14 h, 25°C during dithiothreitol (DTT) and 1 × Complete Proteinase Inhibitor
subjective night). The humidity was 70%. The fluence of light Cocktail (Roche, Mannheim, Germany)]. The beads were
was ∼300 µmol–2 s–1 (400–750 nm) under SD conditions. then washed four times with binding buffer. Bound proteins
were eluted in 1 × SDS sample buffer by boiling for 5 min,
cDNA library and yeast two-hybrid screening separated by 10% SDS–PAGE, transferred to a polyvinylidene
Poly(A)+ RNA was isolated from wild- type N8 rice leaf difluoride (PVDF) membrane (Millipore, Tokyo, Japan) and
blades, harvested at 35–40 d after sowing at zeitberger time subjected to immunoblotting with anti-His antibody (BD
(ZT) 0, 2 and 4, with an RNA Easy Prep kit (TAKARA SHUZO Biosciences Pharmingen). After washing with Tris-buffered
CO. LTD., Kyoto, Japan) and used to construct a cDNA library saline (TBS) containing 0.1% Tween (TBST), the membranes
with a cDNA synthesis kit (Stratagene, La Jolla, CA, USA). were incubated for 1 h with anti-mouse IgG conjugated to
The resultant cDNAs were inserted into the pVP16 vector horseradish peroxidase (GE Healthcare). Detection was per-
and introduced into Saccharomyces cerevisiae L40 cells. A formed using enhanced chemiluminescence (ECL) protein
full-length Hd3a bait was cloned into the pBTM116 vector. gel blot detection reagents (GE Healthcare) and visualized
Independent clones (1.6×106) were screened for interaction using an LAS-1000 Imager (Fujifilm, Tokyo, Japan).

436 Plant Cell Physiol. 50(3): 429–438 (2009) doi:10.1093/pcp/pcp012 © The Author 2009.
 GF14c interacts with the florigen Hd3a in rice

Tissue-specific expression plasmid was introduced at the same time as the marker for
Total RNA from seeds, seedlings, leaf blades, leaf sheaths, transformed cells. The number of cells with reproduced
stems, shoot apices, roots and rice suspension culture cells mVenus fluorescence in the mCherry-expressing cells was
were extracted using TRIZOL reagent (Invitrogen) and scored.
treated with DNase I (Invitrogen). The cDNA was synthe-
sized from 1 µg of total RNA using SuperScript II reverse Supplementary data
transcriptase (Invitrogen). The cDNA was subjected to Supplementary data are available at PCP online.
semi-quantitative analysis of gene expression using PCR with
primers specific for GF14 family members, Hd3a and Funding
ubiquitin (see Supplementary Table S1).
The Ministry of Education, Culture, Sports, Science and
GF14 protein expression in planta Technology, Japan Grants-in-Aid for Scientific Research on
Leaves and stems from wild-type, Hd3a::Hd3a:GFP-overex- Priority Areas (grants 10182102 and 19090013 to K.S.); the

Downloaded from [Link] at University of Windsor on October 19, 2014

pressing, GF14c-overexpressing and gf14c knockout plants Japan Educational Exchanges and Services Iida International
were harvested and ground to a fine powder in liquid nitro- Student Scholarship (to Y.A.P.).
gen. Proteins were extracted as described previously
(Mathieu et al. 2007), using an extraction buffer containing Acknowledgments
150 mM NaCl, 50 mM Tris pH 7.5, 0.1% Tween 20, 10% We thank Shuji Yokoi for help in the early part of this work,
glycerol, 1 mM DTT, 1 mM Pefabloc (Roche), 1× Halt Yukiko Konomi for technical assistance with transgenic
Phosphatase Inhibitor Cocktail (Pierce) and 1× Complete plants, Gynheung An for providing gf14c T-DNA insertion
Proteinase Inhibitor Cocktail (Roche). The extracts were mutants, Seiji Takayama for providing the BiFC vectors, and
centrifuged at 15,000 r.p.m. for 40 min at 4°C. The superna- all members of the Shimamoto lab for helpful discussions.
tant was separated by 12.5% SDS–PAGE and subjected to
immunoblotting using the primary antibody (polyclonal References
rabbit anti-GFP antibody, Abcam) as described previously
(Mathieu et al. 2007) and anti-14-3-3 antibody as described Abe, M., Kobayashi, Y., Yamamoto, S., Daimon, Y., Yamaguchi, A., Ikeda, Y.,
et al. (2005) FD, a bZIP protein mediating signals from the floral
above. After washing with TBST the membranes were incu-
pathway integrator FT at the shoot apex. Science 309: 1052–1056.
bated for 1 h with anti-rabbit and anti-rat IgG conjugated to Ahn, J., Miller, S., Winter, V., Banfield, M., Lee, J., Yoo, S., et al. (2006) A
horseradish peroxidase, respectively (GE Healthcare). Detec- divergent external loop confers antagonistic activity on floral
tion was performed as above. regulators FT and TFL1. EMBO J. 25: 605–614.
Banfield, M. and Brady, R. (2000) The stucture of Antirrhinum
Subcellular localization centroradialis protein (CEN) suggests a role as a kinase regulator.
The GFP sequence was fused to either the C- or the N-terminus J. Mol. Biol. 297: 1159–1170.
of GF14c using the Gateway system (Invitrogen). Expression Chardon, F. and Damerval, C. (2005) Phylogenomic analysis of the
of GF14c-tagged GFP constructs and Hd3a-tagged mCherry PEBP gene family in cereals. J. Mol. Evol. 61: 579–590.
Chen, F., Li, Q., Sun, L. and He, Z. (2006a) The rice 14-3-3 gene family
was driven by the CaMV 35S promoter and the ubiquitin
and its involvement in responses to biotic and abiotic stress.
promoter, respectively. Protoplast isolation from rice Oc
DNA Res. 13: 53–63.
suspension culture (Kyozuka and Shimamoto 1991) and Chen, S., Tao, L., Zeng, L., Vega-Sanchez, M.E., Umemura, K. and
protoplast transfomation (Chen et al. 2006b) were Wang, G.L. (2006b) A highly efficient transient protoplast system for
performed as described previously. After a 24 h incubation analyzing defence gene expression and protein–protein interactions
at 30°C, the protoplasts were examined under a confocal in rice. Mol. Plant Pathol. 7: 417–427.
microscope (LSM510, Zeiss). Cooper, B., Clarke, J., Budworth, P., Kreps, J., Hutchison, D., Park, S., et al.
(2003) A network of rice genes associated with stress response and
Bimolecular fluorescence complementation seed development. Proc. Natl Acad. Sci. USA 100: 4945–4950.
For constructs used in the BiFC experimet, vectors contain- Corbesier, L., Vincent, C., Jang, S., Fornara, F., Fan, Q., Searle, I., et al.
(2007) FT protein movement contributes to long-distance signaling
ing the N- and C-terminal halves of the mVenus were kindly
in floral induction of Arabidopsis. Science 316: 1030–1033.
provided by S. Takayama, NAIST. Hd3a, GF14c and Doi, K., Izawa, T., Fuse, T., Yamanouchi, U., Kubo, T., Shimatani, Z., et al.
β-glucuronidase (GUS)-coding regions were cloned into the (2004) Ehd1, a B-type response regulator in rice, confers short-day
BiFC vectors and purified by the Purelink Plasmid Midiprep promotion of flowering and controls FT-like geneexpression
Kit (Invitrogen). The purified BiFC plasmids were introduced independently of Hd1. Genes Dev. 18: 926–936.
into the rice protoplasts as described previously (Kyozuka Ferl, R. (1996) 14-3-3 proteins and signal transduction. Annu. Rev. Plant
and Shimamoto 1991, Chen et al. 2006b). mCherry expression Physiol. Plant Mol. Biol. 47: 49–73.

Plant Cell Physiol. 50(3): 429–438 (2009) doi:10.1093/pcp/pcp012 © The Author 2009. 437
 Y. A. Purwestri et al.

Giakountis, A. and Coupland, G. (2008) Phloem transport of flowering Monna, L., Lin, H., Kojima, S., Sasaki, T. and Yano, M. (2002) Genetic
signals. Curr. Opin. Plant Biol. 11: 687–694. dissection of a genomic region for a quantitative trait locus, Hd3,
Hanzawa, Y., Money, T. and Bradley, D. (2005) A single amino acid into two loci, Hd3a and Hd3b, controlling heading date in rice.
converts a repressor to an activator of flowering. Proc. Natl Acad. Sci. Theor. Appl. Genet. 104: 772–778.
USA 102: 7748–7753. Muslin, A. and Xing, H. (2000) 14-3-3 proteins: regulation of subcellular
Hayama, R., Yokoi, S., Tamaki, S., Yano, M. and Shimamoto, K. (2003) localization by molecular interference. Cell Signal. 12: 703–709.
Adaptation of photoperiodic control pathways produces short-day Notaguchi, M., Abe, M., Kimura, T., Daimon, Y., Kobayashi, T.,
flowering in rice. Nature 422: 719–722. Yamaguchi, A., et al. (2008) Long-distance, graft-transmissible action
Igarashi, D., Ishida, S., Fukazawa, J. and Takahashi, Y. (2001) 14-3-3 of Arabidopsis FLOWERING LOCUS T protein to promote flowering.
proteins regulate intracellular localization of the bZIP transcriptional Plant Cell Physiol. 49: 1645–1658.
activator RSG. Plant Cell 13: 2483–2497. Park, S., Kim, S., Lee, S., Je, B., Piao, H., Park, S., et al. (2008) Rice
Imaizumi, T. and Kay, S. (2006) Photoperiodic control of flowering: not
Indeterminate 1 (OsId1) is necessary for the expression of Ehd1 (Early
only by coincidence. Trends Plant Sci. 11: 550–558.
heading date 1) regardless of photoperiod. Plant J. 56: 1018–1029.
Ishida, S., Fukazawa, J., Yuasa, T. and Takahashi, Y. (2004) Involvement
Pnueli, L., Gutfinger, T., Hareven, D., Ben-Naim, O., Ron, N., Adir, N.,
of 14-3-3 signaling protein binding in the functional regulation of

Downloaded from [Link] at University of Windsor on October 19, 2014

et al. (2001) Tomato SP-interacting proteins define a conserved
the transcriptional activator REPRESSION OF SHOOT GROWTH by
signaling system that regulates shoot architecture and flowering.
gibberellins. Plant Cell 16: 2641–2651.
Izawa, T., Oikawa, T., Sugiyama, N., Tanisaka, T., Yano, M. and Plant Cell 13: 2687–2702.
Shimamoto, K. (2002) Phytochrome mediates the external light Rice Annotation Project, Tanaka, T., Antonio, BA., Kikuchi, S.,
signal to repress FT orthologs in photoperiodic flowering of rice. Matsumoto, T., Nagamura, Y., Numa, H., et al. (2008) The Rice
Genes Dev. 16: 2006–2020. Annotation Project Database (RAP-DB): 2008 update. Nucleic Acids
Izawa, T. (2007) Daylength measurements by rice plants in Res. 36(Database issue): D1028–D1033.
photoperiodic short-day flowering. Int. Rev. Cytol. 256: 191–222. Roberts, M. (2003) 14-3-3 proteins finds new partners in plant cell
Jaeger, K. and Wigge, P. (2007) FT protein acts as a long-range signal in signaling. Trends Plant Sci. 8: 218–222.
Arabidopsis. Curr. Biol. 17: 1050–1054. Schoonheim, P., Veiga, H., da Costa Pereira, D., Friso, G., van Wijk, K.
Kojima, S., Takahashi, Y., Kobayashi, Y., Monna, L., Sasaki, T., Araki, T., and de Boer, A. (2007) A comprehensive analysis of the 14-3-3
et al. (2002) Hd3a, a rice ortholog of the Arabidopsis FT gene, interactome in barley leaves using a complementary proteomics
promotes transition to flowering downstream of Hd1 under short- and two-hybrid approach. Plant Physiol. 143: 670–683.
day conditions. Plant Cell Physiol. 43: 1096–1105. Tamaki, S., Matsuo, S., Wong, H., Yokoi, S. and Shimamoto, K. (2007)
Komiya, R., Ikegami, A., Tamaki, S., Yokoi, S. and Shimamoto, K. (2008) Hd3a protein is a mobile flowering signal in rice. Science 316:
Hd3a and RFT1 are essential for flowering in rice. Development 135: 1033–1036.
767–774. Tsuji, H., Tamaki, S., Komiya, R. and Shimamoto, K. (2008) Florigen and
Kyozuka, J. and Shimamoto, K. (1991) Transformation and regeneration photoperiodic control of flowering in rice. Rice 1: 25–35.
of rice protoplasts. In Plant Tissue Culture Manual B1. Edited by Wigge, P., Kim, M., Jaeger, K., Busch, W., Schmid, M., Lohmann, J., et al.
Lindsey, K. pp. 1–17. Kluwer Academic Publishers, Dordrecht, The (2005) Integration of spatial and temporal information during floral
Netherlands. induction in Arabidopsis. Science 309: 1056–1059.
Lifschitz, E., Eviatar, T., Rozman, A., Shalit, A., Goldshmidt, A., Amsellem, Z., Wu, C., You, C., Li, C., Long, T., Chen, G., Byrne, M., et al. (2008) RID1,
et al. (2006) The tomato FT ortholog triggers systemic signals that encoding a Cys2/His2-type zinc finger transcription factor, acts as a
regulate growth and flowering and substitute for diverse
master switch from vegetative to floral development in rice. Proc.
environmental stimuli. Proc. Natl Acad. Sci. USA 103: 6398–6403.
Natl Acad. Sci. USA 105: 12915–12920.
Lin, M., Belanger, H., Lee, Y., Varkonyi-Gasic, E., Taoka, K., Miura, E., et al.
Xue, W., Xing, Y., Weng, X., Zhao, Y., Tang, W., Wang, L., et al. (2008)
(2007) FLOWERING LOCUS T protein may act as the long-distance
Natural variation in Ghd7 is an important regulator of heading date
florigenic signal in the Cucurbits. Plant Cell 19: 1488–1506.
and yield potential in rice. Nat. Genet. 40: 761–767.
Lough, T. and Lucas, W. (2006) Integrative plant biology: role of
phloem long-distance macromolecular trafficking. Annu. Rev. Plant Yamamoto, T., Kuboki, Y., Lin, S., Sasaki, T. and Yano, M. (1998) Fine
Biol. 57: 203–232. mapping of quantitative trait loci Hd-1, Hd-2 and Hd-3, controlling
Mathieu, J., Warthmann, N., Kuttner, F. and Schmid, M. (2007) Export heading date of rice, as single Mendelian factors. Theor. Appl. Genet.
of FT protein from phloem companion cells is sufficient for floral 97: 37–44.
induction in Arabidopsis. Curr. Biol. 17: 1055–1060. Yao, Y., Du, Y., Jiang, L. and Liu, J. (2007) Molecular analysis and
Matsubara, K., Yamanouchi, U., Wang, Z., Minobe, Y., Izawa, T. and expression pattern of the 14-3-3 gene family. J. Biochem. Mol. Biol.
Yano, M. (2008) Ehd2, a rice ortholog of the maize INDETERMINATE1 40: 349–357.
gene, promotes flowering by up-regulating Ehd1. Plant Physiol. 148: Yeung, K., Seitz, T., Li, S., Janoschk, P., McFerrank, B., Kaiser, C., et al.
1425–1435. (1999) Suppression of Raf-1 kinase activity and MAP kinase signaling
Mayfield, J., Folta, K., Paul, A. and Ferl, R. (2007) The 14-3-3 proteins mu by RKIP. Nature 401: 173–177.
and upsilon influence transition to flowering and early phytochrome
response. Plant Physiol. 145: 1692–1702. (Received December 26, 2008; Accepted January 21, 2009)

438 Plant Cell Physiol. 50(3): 429–438 (2009) doi:10.1093/pcp/pcp012 © The Author 2009.