Expression Pattern and Subcellular Localization of the Ovate Protein Family in Rice

The Arabidopsis ovate family proteins (AtOFPs) have been shown to function as transcriptional repressors and regulate multiple aspects of plant growth and development. There are 31 genes that encode the full-length OVATE-domain containing proteins in the rice genome. In this study, the gene structure analysis revealed that OsOFPs are intron poor. Phylogenetic analysis suggested that OVATE proteins from rice, Arabidopsis and tomato can be divided into 4 groups (I–IV). Real-time quantitative polymerase chain reaction (RT-qPCR) analysis identified OsOFPs with different tissue-specific expression patterns at all stages of development in the rice plant. Interestingly, nearly half of the total number of OsOFP family was more highly expressed during the seed developmental stage. In addition, seed developmental cis-elements were found in the promoter region of the OsOFPs. Subcellular localization analysis revealed that YFP-OsOFP fusion proteins predominantly localized in the nucleus. Our results suggest that OsOFPs may act as regulatory proteins and play pivotal roles in the growth and development of rice.


Introduction
Development of an organism is based on the temporal and spatial regulation of gene expression, in which transcription factors act as switches of the regulatory cascades [1]. Transcription factor genes comprise a substantial fraction of all eukaryotic genomes [2]. In the Arabidopsis thaliana genome, there are 1922 known and predicted transcription factors, representing >5% of the total number of genes for this genus, distributed across 64 families [2,3]. It has been reported that members of many large transcription factor families play essential roles in plant development [4][5][6][7]. The highest representative in plants is the MADS-box transcription factor family. Members of this family have multiple functions in the transition to flowering, floral organ identity, gametophyte and seed development, fruit development, and lateral root formation [8]. Recently the Arabidopsis ovate family proteins (AtOFPs), forming a plant-specific transcription factor family, were found to control multiple aspects of plant growth and development [9][10][11][12]. There are 18 genes in the Arabidopsis genome that encode proteins with a representation of the distribution of OsOFPs on the rice chromosomes. Although the OsOFPs are scattered throughout the 12 chromosomes of rice (except on chromosomes 6 and 9), their distribution is not uniform (Table 1b). The chromosomal location of the OsOFPs reveals that certain chromosomes and chromosomal regions have a relatively high density of OsOFPs. For instance, there are 8 and 6 OsOFPs located on chromosomes 01 and 05, respectively; however, only a single OsOFP was present on chromosomes 07 and 08, separately (Fig. 1).

Multiple Sequence Alignment and Phylogenetic Analysis of the OFP Proteins
To investigate the phylogenetic relationship among rice, Arabidopsis and tomato OFP genes, the phylogenetic tree was constructed based on their OVATE domain sequences (Fig. 3A). The aligned OVATE domain sequences included 31 OVATE domains from rice, 18 OVATE domains from Arabidopsis and 17 OVATE domains from tomato ( Fig. 3B). As shown in Fig. 3, excluding SlOFP11, all the other OFPs was further divided into 4 major subfamilies, designated I to IV. Except the smallest subfamily IV, each subfamily contains OFP members from rice, Arabidopsis and tomato. However, most of the OsOFP members were clustered in species-specific distinct clades. Only two pairs of orthologs, OsOFP15 and AtOFP09 between rice and Arabidopsis, OsOFP14 and SlOFP22 between rice and tomato could be figured out, comparing with 9 pairs of orthologs presented between Arabidopsis and tomato. These results suggest that the main characteristics of plant OFP proteins in rice, Arabidopsis and tomato were formed before divergence between monocots and dicots, and then evolved separately in a speciesspecific manner.

Expression Patterns of OsOFPs
Gene expression pattern provides important clues for investigating gene function. The tissuespecific expression profiles of the 31 OsOFPs were performed by quantitative real-time RT-PCR (qRT-PCR). Total RNA was isolated from calli, coleoptiles, young leaves, young roots, mature leaves, glumes, young panicles and grains in filling stage (Fig. 4). The results showed that OsOFPs exhibited different expression patterns. The transcript levels of OsOFP01, OsOFP06 and OsOFP29 were higher in the glumes than in other organs (Fig. 4). OsOFP02, OsOFP09, OsOFP10, OsOFP18, OsOFP21 and OsOFP27 were prominently higher in the roots than in other organs (Fig. 4). OsOFP03, OsOFP08, OsOFP15, OsOFP16, OsOFP22, OsOFP25,   OsOFP26 and OsOFP28 were obviously higher in the calli than in other organs (Fig. 4). The expression patterns of OsOFP04, OsOFP05, OsOFP07, OsOFP12, OsOFP13, OsOFP14, OsOFP17, OsOFP20, OsOFP30 and OsOFP31 were all very similar, and they were preferentially expressed in the young panicles tissue (Fig. 4). OsOFP11 and OsOFP24 were more highly expressed in the grains of filling stage but were only minimally expressed in other organs (Fig. 4B, 4C). The levels of OsOFP19 were notably expressed in the coleoptiles (Fig. 4C). OsOFP23 was abundantly expressed in the young leaves but was minimally expressed in other tissues (Fig. 4C). Interestingly,~50% of the total number of OsOFPs was more highly expressed during the seed developmental stage than that during other stages of development. This is similar to the expression profile of the tomato OVATE which is expressed in reproductive organs in early stages of flower and fruit development.
The seed developmental cis-elements that have been well characterized were investigated in 2-kb 5 0 -upstream regions of the OsOFPs [23]. The GCN4 motif is found in the promoter region of the glutelin gene, which encodes a major seed storage protein and is involved in endospermspecific gene expression in rice [24][25][26]. The Skn-1 motif is a consensus sequence in a number of seed-specific promoters and is associated with endosperm-specific gene expression [24]. The promoter region of each OsOFP contains at least one of the 2 seed development cis-elements (S1 Fig.).

Hormone-Induced Expression Profiles of OsOFPs
AtGA20OX1 is thought to be a direct target gene of AtOFP1 and exogenous gibberellic acid can partially rescue the reduced length in rapidly elongating aerial organs of plants overexpressing AtOFP1, suggesting its important role in suppressing GA biosynthesis and maintaining the balance of this hormone in plants [11]. The expression profiles of OsOFPs responding to exogenous GA3 were investigated in our study. Transcript levels of OsOFPs did not change significantly in response to GA treatment (S2 Fig.), indicating that OsOFPs may not participate in the response to GA in rice.
In order to further characterize the relevance between OsOFP family and hormone signaling, the expression patterns of the OsOFP family in response to abscisic acid (ABA) and brassinosteroid (BR), the 2 classes of hormones whose essential function in plant development and stresses resistance were widely appreciated [27][28][29][30][31][32][33], were also assessed using the quantitative real-time RT-PCR (qRT-PCR) in the current study ( Fig. 5; S3 Fig., S4 Fig.). The results showed that only OsOFP29 was upregulated after 24 h of the ABA treatment (Fig. 5A), whereas the transcript levels of other OsOFPs did not change significantly (S3 Fig.). Interestingly, OsOFP03 and OsOFP15 were upregulated immediately under the brassinolide (BL) treatment. The transcript level of OsOFP03 increased quickly and reached its maximum at 2 h (Fig. 5B). OsOFP15 exhibited a response similar to that of OsOFP03 under the BL treatment (Fig. 5C). The quick responses of the OsOFPs to the BL treatment indicate that they may act as important upstream components in BL signal transduction in rice.

Localization of OsOFP Proteins
A previous study has shown that ovate proteins are transcriptional repressors in Arabidopsis [11], but little is known about the molecular function and subcellular localization of OVATE proteins in rice. Usually, transcription regulators are localized in the nucleus to turn a group of target genes on or off. Most of the known AtOFPs are active in the nucleus [9][10][11][12]. The predicted target signal peptides for OVATE proteins in rice were checked using the WoLF PSORT [34] and CELLO [35] prediction programs. According to both programs, most proteins in this family were assigned to either the nucleus or chloroplast (Table 1c, d). Using the WoLF PSORT program, the OsOFPs were predicted to locate in the nucleus or chloroplast (Table 1c). CELLO indicated that most OsOFPs scored relatively higher in the nucleus than in the chloroplast (Table 1d). In total, the differences in the location of OsOFP10, OsOFP12, OsOFP15, OsOFP24, and OsOFP25 were notable between the two prediction programs (Table 1c, d). Although not conclusive, the presented bioinformatics data could serve as a useful reference for further studies on OVATE protein functions in plants.
To confirm the subcellular localization of the OsOFP family, the co-localization of OsOFP yellow fluorescent proteins (YFP) and the nuclear marker AHL22-cyan fluorescent protein [36] were analyzed. The corresponding full-length coding sequences were systematically cloned into a pDONR/Zeo gateway entry vector. This recombined cloning system was subsequently used to further subclone OsOFP coding sequences into the YFP fused destination vectors. As shown in Fig. 6A, 24 OsOFPs were shown to target in the nucleus, and the others were shown to target in the nucleus with granular spots (Fig. 6B). The results were, in general, consistent with the bioinformatics prediction of the subcellular localization of OsOFPs (Table 1c, d, e).

Discussion
OVATE family proteins were present in 13 sequenced plant genomes that represent the major evolutionary lineages of land plants [37]. During the course of plant evolution, most gene families expanded mainly through large scale segmental duplication or tandem duplication to maintain the large number of family members [38], manner of gene duplication had been described in many large transcription factor families such as the bHLH family [39] and the bZIP family [40] in rice. In this study, the relatively high density chromosomal distribution pattern of the OsOFPs indicates that large-scale segmental duplication or tandem duplication may have been involved in the expansion of the OsOFP family. A recent study had verified that segmental duplication and tandem duplication were two major mechanisms for gene expansion in the OsOFP family. For instance, 24 (72.7%) were located within segmental duplication blocks and eight (24.2%) were tandemly duplicated in the rice OsOFP family [37].
Previous studies have shown that OFPs function as transcriptional repressors and regulate multiple aspects of plant growth and development [9,[11][12][13]41]. The OVATE gene determines the conversion of the tomato fruit from round-to pear-shaped, and overexpression of this gene results in smaller-sized floral organs and leaflets [13]. The tomato OVATE transcripts can be detected in flowers 10 days before anthesis and until 8 days after anthesis in developing fruit [13]. Recently, Huang et al. (2013) reported that the OVATE expression increased in fruit at 33 days post of anthesis, suggesting its important role at the fruit ripening stage [15]. To some extent, there is coincidence between the spatially restricted expression of OVATE and the pearshaped fruit shape phenotype [13]. Round seeds or blund-end siliques were also observed in AtOFP family members overexpressed A. thaliana respectively [11,12]. Furthermore, an Ovate-like gene in pepper was reported to be involved in determining fruit shape [41]. In this study, tissue-specific expression pattern analysis showed that OsOFPs were expressed in all developmental phases of the rice plant, and nearly half of this gene family was more highly expressed during the seed developmental stage (Fig. 4). Moreover, each OsOFP contained the GCN4 or Skn-1 motifs, 2 of the seed developmental cis-elements that have been well characterized in rice, in the promoter region (S1 Fig.). Results from the functional analysis of the OFPs in the above plant species and from the current study hint at key roles of OsOFPs in controlling important traits related to seed development in rice.
TFs function as key regulators during different stages of seed development [42][43][44]. A pair of basic helix-loop-helix (bHLH) proteins, PGL1 and APG, are involved in determining grain weight and length antagonistically, in which APG is a negative regulator whose function is inhibited by PGL1 [42]. The homeobox family of genes was found to play a pivotal role throughout the process of seed development. For instance, homeobox gene OSH1 is critical for the regionalization of cell identity during early embryogenesis [43]. Studies in Arabidopsis showed that AtOFPs interact with TALE homeodomain proteins to regulate plant development [9]. Results of most OsOFPs with stronger expression profiles in the mature seed or embryo suggest that OsOFPs may act as important transcriptional regulators during seed development in rice. Further studies are needed to identify the interactions of OsOFPs in order to discover new seed development regulating pathways that include OsOFP at their core.
Many transcriptional factor families were characterized to play important roles in different developmental processes regulated by phytohormones [45][46][47]. So far, only AtOFP1 was reported to target AtGA20ox1, a key enzyme gene in gibberellin biosynthesis, and to suppress its expression in plants overexpressing AtOFP1 [9,11]. No information about the relevance between plant OFP families and the other phytohormones was provided. In this study, the expression profiles of OsOFPs responding to exogenous GA3, ABA and BL were investigated ( Fig. 5;  S2 Fig., S3 Fig., S4 Fig.). We found that the transcript level of OsOFP15 increased rapidly in response to BL treatment, meanwhile it was specifically expressed in the mature seed. Brassinolide is the most bioactive form of the growth-promoting plant steroid hormones, termed brassinosteroids. Brassinosteroids are essential for a wide range of developmental and physiological processes [28][29][30][31]. BR-deficient and BR-insensitive mutants display dwarfism, erect leaves, and altered panicle length in combination with a reduced seed setting [48][49][50]. XIAO, a putative LRR receptor-like kinase, is related to BR biosynthesis; and the loss-of-function mutant, xiao, leads to reduced plant size and seed setting due to a decrease in the rate of cell division [49]. Overexpression of the BR biosynthesis related gene Zm-CYP-1 increased grain yield by controlling the seed filling in rice [31]. The spatially restricted and BR-induced expression pattern of OsOFP15 suggest that it may participate in regulating the BR response in a rice seed setting.
Subcellular proteomics are helpful in exploring the biological functions of proteins. In the current study, the subcellular localization analysis clearly revealed that the YFP-OsOFP fusion proteins predominantly localized in the nucleus. With regard to OsOFP05, 28, and 30, only very weak yellow fluorescent signals were observed in the nucleus (Fig. 6), while strong yellow fluorescent signals were clearly observed in the nucleus for other OsOFPs (Fig. 6). The reason for these observations may be that OsOFP05, 28, and 30 were not stable when compared to the other OsOFPs in the nucleus of Arabidopsis protoplast cells. We suspected that the N-terminus-fused YFP may be the reason for the instability of the OsOFPs with weak fluorescent signals. As exhibited in Fig. 6, OsOFP02, 03, 04, 05, 06, 10, 15, and 17 prominently displayed nuclear concentrations with multiple big and small granular spots floating along the nucleus. Some transcription factor proteins were nuclear localization and also present in the cytoplasm [51]. For instance, the KNOX proteins, encoding transcription factors involved in shoot apical meristem development and maintenance, had similar subcellular distribution patterns to these OsOFP proteins [52]. BZR1 and BZR2/BES1, the two key transcription factors of the BR signaling pathway, were found in both the nucleus and cytoplasm [53,54]. BR signal increased the nuclear accumulation of BZR1 and BZR2/BES1 [53][54][55]. The subcellular distrbution patterns of these OsOFPs suggested that they may function as potential nucleocytoplasmic shuttling proteins and can be regulated by various signals such as hormone and light [56].
The localization prediction programs for OsOFPs on the Website do not support the actual subcellular localization for some proteins. For example, OsOFP07, OsOFP10, OsOFP12, OsOFP14, OsOFP15, OsOFP24, OsOFP25, OsOFP26 and OsOFP28 did not localize in the nucleus according to the predictions of the WoLF PSORT or CELLO programs (Table 1c, d, e). However, OsOFPs were shown to target to the nucleus in the current study. This discrepancy may be explained by the fact that the nuclear localization signals identified and deposited in the WoLF PSORT and CELLO databases are not complete. It is also possible that novel nuclear localization signals, that have not yet been identified, could exist in the OsOFP sequences.
In conclusion, this study provided the genomic framework, intron number, and phylogenetic analysis of the 31 OsOFPs. In addition, we studied the expression profile of the whole family during the entire life cycle of the rice plant and under GA, ABA, and BL treatments. Furthermore, the subcellular localization of OsOFPs was also investigated. Our data provide insight into the roles of OsOFPs during the seed developmental stages and in response to BL. This study provides a useful reference for conducting more detailed functional analyses of these OFPs in rice and will be helpful in the selection of appropriate candidate genes for further studies.

Phylogenetic analysis and sequence alignment
Multiple sequence alignment were performed using the ClustalX version 1.83 [57] and were manually corrected. The obtained sequence alignments were used as input to construct phylogenetic tree with the neighbor-joining algorithm in MEGA 6.06 [58]. Bootstrap analysis was performed using 1000 replicates.

Plant Material and Growth Conditions
Wild-type Nipponbare (O. sativa L. ssp. japonica) was used in the current study. The seeds were imbibed for 48 h at 30°C and cultivated in the test field or growth chamber under diurnal conditions (day phase: 10 h/30°C; night phase: 14 h/24°C). For tissue-specific expression analysis, materials were collected at different developmental stages. Young leaves, young roots, calli and coleoptiles were collected from the growth chamber; and mature leaves, glumes, young panicles and grains in filling stage were collected from the test field of Jilin University. All plant materials were collected in liquid nitrogen and stored at -80°C for RNA extraction.
Arabidopsis seeds of the Columbia ecotype (Col-0) were used in our study. Seeds were surface-sterilized, vernalized, and sown and grown on 1/2 MS media until the leaves reached the four-leaf stage. Then, the seedlings were transplanted to a growth chamber containing peat moss mixed with vermiculite (1/1, v/v), under diurnal conditions (day phase: 8 h/22°C; night phase: 16 h/18°C). The relative humidity was maintained at 60-70%. Leaves were collected from 3-to 5-week-old seedlings for protoplast isolation and transfection assays.

RNA Isolation and cDNA Synthesis
Total RNA was extracted using the TRIzol reagent (Invitrogen, Carlsbad, CA, USA) according to the manufacturer's instructions. The integrity of the RNA was checked electrophoretically, and complementary DNA (cDNA) was synthesized using 4 μg of RNA with the Oligo d(T) 18 primers and M-MLV reverse transcriptase (TaKaRa Bio, Tokyo, Japan), according to the manufacturer's protocol.

Real-time Quantitative PCR Analysis
Relative transcript level expression profiles of OsOFPs were evaluated by real-time quantitative PCR (RT-qPCR) on Mx3005P (Stratagene, La Jolla, CA, USA) using the SYBR Premix ExTaq polymerase (TaKaRa, Bio Inc.). Each reaction contained 12.5 μl of the 2× SYBR Premix ExTaq, 50-100 ng of the cDNA template, 0.5 μl of 10 mM of each primer, and 10.5 μl of doubledistilled H 2 O for a final volume of 25 μl. The PCR reaction parameters were 95°C for 30 s (1 cycle), 95°C for 5 s, and 60°C for 20 s (40 cycles), which was followed by a melting curve analysis at 95°C for 60 s, 55°C for 30 s, and 95°C for 30 s. The relative fold differences were calculated based on the comparative 2 -ΔΔ Ct method. Approximately 100-200-bp PCR products, unique to each OsOFP, were amplified; the housekeeping gene ACTIN (X15863.1) was used to normalize the transcript level of each OsOFP in the samples. Expression profile analysis of OsOFP genes in each type tissue was displayed using relative percentage [15] in this study. The specific primer pairs used were listed in S3 Table. Expression Profiles of OsOFPs in Response to ABA and BL The rice seeds of the wild-type Nipponbare were sterilized with 3% NaOCl for 30 min, washed extensively with distilled water, and placed in petri dishes with wetted filter papers at 30°C in the dark. After 3 d of incubation, germinated seeds were sown in a 96-well plate, with the bottom removed, and the plate was fixed in a 1/2 MS culture solution at 26°C under continuous light. After 14 d, the seedlings were subjected to ABA and BL treatments. For the ABA or BL treatments, seedlings were transferred to a culture solution with 50 μM of ABA, 1 μM of BL, or without ABA and BL (i.e., the control). Subsequently, young leaves were collected at 0, 2, 4, 8, 16, 24, 36, and 48 h after treatment. Then the expression profiles of OsOFPs responding to ABA and BL were evaluated by real-time quantitative PCR (RT-qPCR).

Construction of the Subcellullar Localization Plasmid
The expression vectors of subcellullar localization were constructed as follows: the coding sequences of the OsOFPs were cloned into an entry vector (pDONR/Zeo; Invitrogen) by using the BP-clonase, according to the manufacturer's instructions, and subsequently cloned into the destination vector (pENSG-YFP) of the N-terminus fused the YFP reporter gene by an LR reaction (Gateway recombination, Invitrogen). All of the fusion constructs were driven by the 35S promoter. The recombinant plasmids vector pENSG-CFP-AHL22, carrying a nuclear marker gene AHL22 [36], was co-transformed with each target gene. The specific primer pairs used were listed in S4 Table. Protoplast Isolation, Transformation, and Confocal Microscopy Leaves of 3-to 5-week-old plants were used in the subcellular localization analysis. The leaves were cut into 0.5-1-mm strips with fresh razor blades (without wounding) and incubated in an enzyme solution that included 1% cellulase R10, 0.25% macerozyme R10, 0.4 M mannitol, 20 mM KCl, 10 mM CaCl 2 , and 20 mM MES for 3-4 h, with 40-50 rpm slow shaking. After incubation, the protoplast suspension was filtered through a metal sieve and centrifuged at 100 × g for 5 min. The pelleted protoplasts were suspended in 5 mL of a W5 solution (154 mM NaCl, 125 mM CaCl 2 , 5 mM KCl, and 2 mM MES/KOH; pH 5.7) and centrifuged for 5 min at 100 × g. The protoplasts were transferred to a tube containing 5 mL of the W5 solution. The protoplasts were pelleted again by centrifugation at 100 × g for 5 min and resuspended in 5 mL of the W5 solution. The protoplasts were incubated on ice for 30 min. The protoplasts were again resuspended in 5 mL of MMg buffer (400 mM mannitol, 15 mM MgCl 2 , and 4 mM MES/ KOH; pH 5.7).
For cotransformation, 10 μg of each plasmid DNA was added to 100 μL of the protoplast suspension. An equal volume of 40% (w/v) PEG3350, freshly prepared with 0.1 M CaCl 2 and 0.8 M mannitol solution, was added. Then, the mixture was incubated at room temperature for 30 min. After incubation, the mixture was diluted with 500 μL of the W5 solution. The solution was fully mixed, and the protoplasts were pelleted by centrifugation at 100 × g for 5 min. Then, the protoplasts were washed twice using the W5 solution, resuspended gently in 1 mL of the W5 solution, and incubated in 12-well plates at room temperature for 18-20 h in darkness.
All microscopic observations were performed using a confocal laser scanning microscope. The fluorescence of the YFP was visualized with excitation and emission wavelengths of 488 and 505-530 nm, respectively. For CFP, the excitation and emission wavelengths were 458 and 465-530 nm, respectively. Chloroplast autofluorescence was visualized in a detection channel with excitation and emission wavelengths of 488 and 650-710 nm, respectively. Image processing was performed with ImageJ (http://rsb.info.nih.gov/ij/).