•The growth of a plant organ to its characteristic size is regulated by an elaborate developmental program involving both internal and external signals. Here, we identify a novel Arabidopsis gene, ORGAN SIZE RELATED1 (OSR1), that is involved in regulation of organ growth and overall organ size.
•A combination of genetic, cytological and molecular approaches was used to characterize the expression profile, subcellular localization and roles of OSR1 during organ growth.
•Ectopic expression of OSR1 in Arabidopsis resulted in enlarged organs, as a consequence of increases in both cell number and cell size. OSR1 shares a conserved OSR domain with ARGOS and ARGOS-LIKE (ARL), which is sufficient for their functions in promoting organ growth. OSR1 is a plant hormone-responsive gene and appears to act redundantly with ARGOS and ARL during organ growth. The OSR proteins are localized to the endoplasmic reticulum.
•Our results suggest that three co-evolved members of the OSR family may act coordinately to orchestrate growth signals and cell proliferation and expansion, thereby affecting organ growth and final organ size.
The architecture of a plant depends greatly on the number, size and shape of its organs, and organ size is undoubtedly of great importance to agronomic yield and biomass production in plants (Gonzalez et al., 2009). The relative constancy of organ size within a given species, but its remarkable variance among species, suggests that the final size of an organ is defined by intrinsically developmental programs (Mizukami, 2001). Although the key pathways that limit organ growth in insects and mammals have been characterized (Dong et al., 2007), the counterparts of many factors that determine organ size in insects and mammals are not found in plants, suggesting that the mechanism underlying organ size control in plants differs from that in animals. Given their sessile and post-embryonic organogenesis lifestyle, organ growth in plants is also greatly influenced by internal and environmental signals, including plant hormones, light, temperature and nutrients, and the final size of organs is largely influenced by the combined effect of these signals during organogenesis (Tsukaya, 2003, 2005).
In plants, the growth of an organ to its characteristic size occurs in two successive but overlapping phases: the cell proliferation phase, during which cell number increases dramatically, and the subsequent cell expansion phase, during which cells expand to appropriate sizes (Mizukami, 2001; Anastasiou & Lenhard, 2007). These two phases are referred to as ‘growth by cell proliferation’ and ‘growth by cell expansion’, respectively (Anastasiou & Lenhard, 2007). In Arabidopsis leaves, most cells begin to undergo cell expansion when cell proliferation arrests gradually from leaf tip to base; however, there are still some cells that maintain the meristematic competence to continue dividing and form the specific cell type within each cell layer (Donnelly et al., 1999), suggesting that cell proliferation and expansion are coordinately controlled in development (Tsukaya & Beemster, 2006).
Recent genetic analyses have identified several factors that control organ size by regulating the cell proliferation and/or cell expansion process in plants. Some of these factors have roles in positively regulating cell proliferation during organ growth, such as AINTEGUMENTA (ANT), ANGUSTIFOLIA 3 (AN3), GROWTH-REGULATING FACTOR 5 (AtGRF5), JAGGED (JAG), STRUWWELPETER (SWP), SWELLMAP1 (SMP1) and KLUH (KLU). Ectopic expression of these genes in Arabidopsis prolongs the duration of cell proliferation, leading to the production of larger organs with more cells, while mutations in the genes encoding these factors reduce the duration of cell proliferation and thus result in smaller organs (Mizukami & Fischer, 2000; Autran et al., 2002; Dinneny et al., 2004; Ohno et al., 2004; Clay & Nelson, 2005; Horiguchi et al., 2005; Anastasiou et al., 2007; Lee et al., 2009). By contrast, some other factors, including AUXIN RESPONSE FACTOR2 (ARF2), BLADE ON PETIOLE1 (BOP1), PEAPOD1/2 (PPDs), BIG BROTHER (BB) and DA1, appear to restrict organ growth by limiting the period of proliferation, because loss of function in each of these genes results in enlarged organs as a result of the increased cell number (Ha et al., 2003; Disch et al., 2006; Schruff et al., 2006; White, 2006; Li et al., 2008). A general theme emerging from these studies is that the regulation of cell proliferation time, rather than cell division rate, is one of critical mechanisms for determining final organ size in plants (Anastasiou & Lenhard, 2007). However, a few factors that regulate organ growth by cell expansion have also been identified. For example, ROTUNDIFOLIA3 (ROT3) stimulates polar cell expansion in the longitudinal direction in Arabidopsis leaves (Kim et al., 1998), and overexpression of AtGRF1/2 increases leaf size by generally increasing cell size (Kim et al., 2003). BIGPETAL, a putative transcription factor in Arabidopsis, keeps petal size in check by limiting cell expansion (Szécsi et al., 2006).
The Arabidopsis gene ARGOS was identified as an auxin-inducible gene that encodes a small plant-specific protein without a known functional domain. Overexpression of ARGOS in Arabidopsis dramatically increases lateral organ size through prolonged expression of the APETALA2 (AP2) transcription factor gene ANT, which affects cell proliferation by regulating the expression of Cyclin D3;1 (CycD3;1) during organ growth (Krizek, 1999; Mizukami & Fischer, 2000; Hu et al., 2003). In the Arabidopsis genome, there is a homologue of ARGOS, named ARGOS-LIKE (ARL), and it is interesting that ARL appears to regulate organ growth mainly by influencing cell expansion rather than cell proliferation (Hu et al., 2006). A recent study revealed that there is a single-copy orthologue of ARGOS in the rice (Oryza sativa) genome, and detailed cellular analysis of transgenic Arabidopsis plants overexpressing OsARGOS indicated that OsARGOS affects organ growth by promoting both cell proliferation and cell expansion (Wang et al., 2009). In addition, ARGOS and OsARGOS are induced by the phytohormones auxin and cytokinin, while ARL is induced by brassinosteroid (BR) (Hu et al., 2003, 2006; Wang et al., 2009), suggesting that ARGOS/ARL-modulated organ growth also involves environmental or growth cues during plant development.
To gain insights into ARGOS/ARL-regulated organ growth, we attempted to identify other factors involved in the ARGOS/ARL-related pathway. Here, we report a novel Arabidopsis gene, ORGAN SIZE RELATED1 (OSR1), that regulates organ growth and final organ size by affecting both cell proliferation and expansion. OSR1 shares a small conserved OSR motif with ARGOS and ARL, which is required and sufficient for promoting organ growth. The three OSR genes show different responses to plant hormones and appear to be co-regulated during organ development. Our data suggest that co-evolved members of the OSR family may act coordinately in response to growth signals to regulate cell proliferation/expansion during plant organogenesis.
Materials and Methods
Plant materials and growth conditions
The Arabidopsis thaliana (L.) Heynh Columbia-0 accession was used in this study. Sterilized seeds were plated on 1/2 MS medium containing 1% sucrose and 0.8% agar, and vernalized at 4°C in darkness for 2–4 d. The plate was then transferred to a culture room at 22 ± 1°C with illumination of 80–90 μmol m−2 s−1 with a 16-h light : 8-h dark photoperiod for seed germination. For hormone treatment or root observation, seedlings were grown vertically for 7–12 d. For morphological analysis, 7-d-old seedlings were transferred to soil and grown in a growth room at 22 ± 1°C with the same illumination and photoperiod as in the culture room (Jing et al., 2009).
The 267-bp OSR1 coding sequence was amplified by reverse transcription–polymerase chain reaction (RT-PCR) and cloned into pVIP96 for generation of the 35S-OSR1 construct (Hu et al., 2003). To generate truncated OSR genes, cDNA fragments of OSR1, ARGOS or ARL with an added ATG and/or a stop codon were amplified and cloned into pVIP96. To monitor tissue-specific expression of OSR1, a 1.9-kb promoter region and the 433-bp 3′-untranslated region of OSR1 were fused with the β-glucuronidase (GUS) gene in pBI101 to form the proOSR1-GUS construct. The coding sequence of OSR1, ARGOS or ARL was fused with the GFP gene and cloned into pVIP96 to form the 35S-OSR1-GFP, 35S-ARGOS-GFP or 35S-ARL-GFP construct, respectively. All transgenic plants were generated by Agrobacterium tumefaciens-mediated transformation (Zhang et al., 2006). Approximately 20 independently transgenic lines for each construct were generated and at least three T3 homozygous lines with a single T-DNA insertion were used for detailed analyses.
Morphological and cytological analyses
The leaf and petal were used as representative organs to determine organ size and to assess the contributions of cell number and cell size. Fully expanded fifth leaves were excised and photographed, and then visualized under a microscope after clearing with chloral hydrate, as previously described (Jing et al., 2009). The palisade cells at the central region of a half leaf beside the mid-vein were photographed to determine cell size and cell number per unit area. To determine petal size and to examine epidermal cells, fully expanded petals were excised and photographed, and then fixed in fresh Formaldehyde-Acetic Acid-Alcohol (FAA) solution and scanned with a scanning electron microscope (SEM; Hitachi S-4800; Hitachi, Tokyo, Japan). The areas of leaf blades, petals and cells were measured with image j software (http://rsbweb.nih.gov/ij/). The total number of palisade cells per leaf or epidermal cells per petal was calculated as the product of the leaf area or petal area and the cell number per unit area was determined using the microscope (Hu et al., 2010).
Gene expression analysis
Total RNA was isolated with a guanidine thiocyanate extraction buffer, and RNA gel blots were performed as described previously (Hu et al., 2000). Real-time quantitative RT-PCR (qRT-PCR) was carried out using a Rotor-Gene 3000 thermocycler (Corbett Research, Sydney, Australia) with the SYBR® Premix Ex Taq™ II kit (Takara, Dalian, China). The normalization and relative values of expression level for each gene were calculated from three biological replicates, as previously described (Hu et al., 2010). The primers used were as follows: for ANT, 5′-AAGCACGGATTGGTAGAGTCG-3′ and 5′-GCATTTGTGCCACGGAACTTA-3′; for CycD3;1, 5′-GCAAGTTGATCCCTTTGACC-3′ and 5′-CAGCTTGGACTGTTCAACGA-3′; and for the internal control ACTIN2, 5′-GCTCCTCTTAACCCAAAGGC-3′ and 5′-CACACCATCACCAGAATCCAGC-3′. For the GUS staining assay, seedlings or organs of homozygous transgenic plants were incubated in a 50 mM sodium phosphate solution (pH 7.0) containing 5 mM K4Fe(CN)6, 5 mM K3Fe(CN)6, 0.1% Triton X-100, and 1 mM 5-bromo-4-chloro-3-indolyl-β-glucuronic acid (Gluc) at 37°C for several hours.
Protoplast preparation and co-transformation
The OSR1-, ARGOS- and ARL-GFP fusion genes cloned in the pUC vector were used for co-transformation with the plasmid containing an endoplasmic reticulum (ER)-localized chaperon-binding protein-RFP (Bip-RFP) (Kim et al., 2001). Arabidopsis protoplasts were prepared and the fusion constructs were introduced into protoplasts by polyethylene glycol-mediated transformation according to the method described by Jin et al. (2001). The co-transformed protoplasts were incubated at 22°C in the dark for c. 12 h before visualization of GFP and RFP fluorescence under a confocal microscope.
Yeast two-hybrid (Y2H) assay
The coding regions of OSR1, ARGOS and ARL were cloned into pGADT7 and pGBKT7 to fuse with the activation domain (AD) and DNA-binding domain (BD), respectively. The constructs were then transformed into Saccharomyces cerevisiae strain AH109 (BD, Biosciences, Palo Alto, CA, USA) according to the manufacturer’s protocol. Self-activation for each single construct was not detectable when yeast cells were cultured on SD selective medium (SD-His-Leu + 5 mM 3-AT or SD-His-Trp + 5 mM 3-AT). Co-transformation was conducted with each AD fusion construct and BD fusion construct, to test whether each OSR can interact with itself and other OSRs in yeast cells. The co-transformed yeast cells with OsMADS7-AD and OsMADS13-BD were used as a positive control (Cui et al., 2010).
Phylogenetic tree construction
To elucidate the phylogenetic relationships among OSR-like genes, 39 predicted OSR homologues from 14 plant species were identified in the National Center for Biotechnology Information (NCBI) database and Phytozome (http://www.phytozome.net), and the coding sequences were used to construct a phylogenetic tree. All these OSR-like coding sequences were first aligned using muscle 3.6 (Edgar, 2004), and local alignment was adjusted manually using GeneDoc (Nicholas et al., 1997). PhyML version 2.4 was used to construct the phylogenetic tree and the parameters were set as described by Guindon & Gascuel (2003). The coding sequence of OSR homologue Physcomitrella patens 169695 was used as an outgroup for the phylogenetic tree. The reliability of each interior branch of the tree was assessed by bootstrapping with 1000 replications.
The expression profile of OSR1 is similar to that of ARGOS
Our previous work showed that ARGOS and its homologue ARL regulate lateral organ growth mainly by affecting cell proliferation and cell expansion, respectively (Hu et al., 2003, 2006). To further expand our knowledge of ARGOS/ARL-mediated organ development, we utilized the Arabidopsis thaliana trans-factor and cis-element prediction database (ATTED II) to identify the candidate genes that were likely to be co-expressed or co-regulated with ARGOS/ARL (http://atted.jp/). In this database, an Arabidopsis gene, At2g41230, was shown to be directly connected to both ARGOS and ARL in a co-expression network (Fig. 1a). We then compared the expression profiles of At2g41230, ARGOS and ARL during development using the AtGenExpress atlas, which comprises the expression profiles of 22 746 probe sets on an Affymetrix ATH1 array with triplicate expression estimates from 79 diverse development samples, ranging from embryogenesis to senescence and from roots to flowers (Schmid et al., 2005). The expression profiles generated by the AtGenExpress atlas showed that the expression profiles of At2g41230 and ARGOS were indeed comparable in most developing organs and stages, except in the apex and in flowers. The expression pattern and signal intensity, in contrast, were quite different between At2g41230 and ARL (Supporting Information Fig. S1a). The expression profiles generated by Electronic Fluorescent Pictograph (e-FP) (Winter et al., 2007) also suggested that At2g41230 and ARGOS were expressed quite similarly in various organs (Fig. S1b).
The publicly available cDNA of At2g41230 is 749 bp long, and encodes a predicted novel protein of 88 amino acids (Fig. 1b). We named At2g41230 ORGAN SIZE RELATED1 (OSR1). Surprisingly, a detailed protein alignment revealed that OSR1 appeared to contain a motif that is conserved in ARGOS and ARL (Fig. 1b).
To further determine the tissue-specific expression of OSR1, we generated transgenic Arabidopsis plants carrying an OSR1 promoter::β-glucuronidase (proOSR1-GUS) fusion gene and examined GUS activities in transgenic plants. In 7-d-old seedlings, abundant GUS staining was detected in leaf primordia, and low GUS expression was detected in cotyledons (Fig. 1c). In expanding leaves, GUS staining was still detectable in the blade and the petiole, especially at the dispersed meristematic regions and the leaf margin (Fig. 1d). High GUS expression was also observed in floral organs (Fig. 1e).
Ectopic expression of OSR1 increases final organ size predominantly by enhancing cell proliferation
To test whether OSR1 has a function in the regulation of organ size, as does ARGOS, we generated transgenic 35S-OSR1 plants and examined their organ development. Compared with control lines, which contained an empty vector, all of 20 independently transgenic lines overexpressing OSR1 displayed visibly enlarged aerial organs, including leaves, cotyledons, and floral organs (Fig. 2a,b, Table 1). As a result, the heights of transgenic plants were also increased (Table 1). RNA gel blot analysis of three independent homozygous T3 lines confirmed the high expression levels of the OSR1 transgene in these transgenic plants (Fig. 2c). Detailed characterization of the fifth rosette leaf and the petal showed that, although the shape of both the leaf and the petal in 35S-OSR1 transgenic plants was not altered, the average areas of the leaf blade and the petal in 35S-OSR1 increased by c. 80% when compared with those in control plants (Fig. 2b,d). Moreover, overexpression of OSR1 resulted in delayed flowering and longer roots (Table 1). These observations indicate that OSR1 has a role in regulation of organ growth and final organ size.
Table 1. Phenotype of 35S-ORGAN SIZE RELATED1 (OSR1) transgenic plants
Student’s t-test: *, P < 0.05; **, P < 0.01.
Cotyledon area (mm2)
2.66 ± 0.37 (n = 10)
4.21 ± 0.48 (n = 10)**
Root length (mm)
24.4 ± 2.1 (n = 20)
30 ± 2.9 (n = 20)*
Flowering time (d)
28.1 ± 1.1 (n = 16)
34.3 ± 1.3 (n = 16)**
Silique length (mm)
12.6 ± 0.7 (n = 20)
16.3 ± 1.5 (n = 20)*
Plant height (cm)
32.6 ± 1.9 (n = 8)
42.1 ± 3.2 (n = 8)**
Further comparison of the growth kinetics of the fifth leaves between control and 35S-OSR1 plants showed that the leaf blades expanded at a similar rate from emergence to 5 d, and the size difference in the two genotypes was produced mainly afterwards (Fig. 2e), suggesting that the size increase in the 35S-OSR1 leaves may result from prolonged cell expansion and/or cell proliferation. To assess the contributions of cell proliferation and expansion to the increased organ size in 35S-OSR1 plants, we first visualized the palisade cells of fully expanded fifth leaves in the two genotypes. As shown in Fig. 3(a,b), the average size of palisade cells in 35S-OSR1 increased by c. 20% compared with that of controls, while the estimated palisade cell number per leaf increased by > 50%, demonstrating that the enlarged leaf in 35S-OSR1 is a result of increases in both cell number and size. Similarly, the increase in petal size of 35S-OSR1 could also be attributed to increases in both cell number and cell size, with cell number contributing c. 60% of the enlargement (Fig. 3c,d). To further substantiate the predominant role of OSR1 in cell proliferation, we also introduced proCycB1;1-Dbox-GUS into 35S-OSR1 plants. As the CycB1;1-GUS reporter marks the cells at the G2-M phase (Criqui et al., 2001), the enhanced GUS activities observed in young leaves and the root meristem region of 35S-OSR1 seedlings indicated the presence of more proliferating cells in the transgenic plants than in the control plants (Fig. S2). Therefore, we conclude that the size increase in 35S-OSR1 organs results from increases in both cell proliferation and cell expansion, with cell proliferation as the dominant factor.
OSR1 modulates cell proliferation through the expression of ANT
As 35S-OSR1 and 35S-ARGOS plants are similar in terms of their organs being enlarged primarily through enhanced growth by cell proliferation, we suspected that OSR1 modulates cell proliferation through the regulation of ANT, as does ARGOS. To test this hypothesis, we first compared the expression of ANT and CycD3;1 in both expanding and expanded leaves between control and 35S-OSR1 plants. Similar to those observed in 35S-ARGOS plants, the expression levels of ANT and CycD3;1 in expanding leaves were comparable between the two genotypes. However, persistent expression of ANT and CycD3;1 was detected even in leaves of 40-d-old 35S-OSR1 plants (Fig. 4a). This finding demonstrates that overexpression of OSR1 prolonged the expression of ANT and CycD3;1, and thus the duration of cell proliferation, during leaf development. We then introduced the ant-1 mutation into 35S-OSR1 transgenic plants, and, as expected, the ant-1 mutation dramatically blocked leaf enlargement in 35S-OSR1 plants (Fig. 4b). Because ANT stimulates organ growth by promoting cell proliferation (Mizukami & Fischer, 2000), we reasoned that the slight size increase in 35S-OSR1/ant-1 leaves may be attributable to the effect of OSR1 on cell expansion. To test this hypothesis, we carefully determined palisade cell size and the number of fully expanded fifth leaves in ant-1 and 35S-OSR1/ant-1 plants, and found that the ant-1 mutation indeed completely blocked the effect of 35S-OSR1 on cell proliferation, but not its effect on cell expansion (Fig. 4c,d), confirming that OSR1-mediated cell proliferation, but not cell expansion, occurs through the ANT pathway during organogenesis.
The OSR domain is sufficient to promote organ growth
As OSR1 shares a conserved domain with ARGOS and ARL, which comprises an identical LPPLPPPP motif and two putative transmembrane helices (Fig. 5a), we named it the Organ Size Related (OSR) domain. It is likely that the OSR domain is responsible for the function of the three members of the OSR family. To test whether this is the case, we generated transgenic plants overexpressing different truncated OSR1 coding regions and examined their final leaf sizes. As shown in Fig. 5(b,c), transgenic plants harbouring a transgene that encodes a truncated OSR1 protein with an intact OSR domain or the OSR domain alone still exhibited enlarged organs, as did 35S-OSR1 plants. By contrast, overexpression of a transgene with a disrupted OSR domain could not recapitulate the organ phenotype of 35S-OSR1 plants. Consistently, overexpression of the OSR domain in ARGOS and ARL also resulted in the organ phenotypes of 35S-ARGOS and 35S-ARL, respectively (Fig. 5b,c). These results demonstrate that the OSR domain is essential and sufficient for promoting organ growth.
OSR proteins are ER-localized
To elucidate the biological roles of OSR proteins, we first visualized subcellular localization of the OSR1 protein in transgenic plants carrying 35S-OSR1-GFP. The increased organ size in the transgenic plants indicated that the fusion gene was functional (Fig. S3a). The GFP florescence signals observed in root cells before and after plasmolysis suggested that OSR1 was localized in the cytoplasm (Fig. 6a). Because there are two predicted transmembrane helices in OSR1, we suspected that they may be localized to the endomembrane system, probably to the ER. We then investigated whether the OSR1-GFP fusion protein co-localized with the ER marker protein Bip-RFP in protoplasts. As shown in Fig. 6(b), the green fluorescence signals of OSR1-GFP fusion proteins closely overlapped the red fluorescence signals of Bip-RFP, indicating that OSR1 is localized to the ER. Similarly, the other two OSR proteins, ARGOS and ARL, were also found to be ER-localized (Fig. S3a–e).
OSR1 regulates organ growth redundantly with ARGOS and ARL
To further investigate the relationships among OSR1, ARGOS and ARL in organogenesis, we obtained the T-DNA insertion mutant osr1-1 (GABI_436G04) from Nottingham Arabidopsis Stock Centre (NASC) and argos-1 (SAIL_896_G10) from Arabidopsis Biological Resource Center (ABRC), in which a T-DNA fragment was inserted into the OSR1 coding region or the ARGOS 5′-untranslated region, respectively (Fig. S4a). RNA blot analysis revealed that osr1-1 and argos-1 were knock-out mutants (Fig. S4b). Neither osr1-1 nor argos-1 displayed an obvious reduction in organ size compared with wild-type plants. However, the osr1-1 argos-1 double mutant displayed smaller organs, with a c. 12% reduction in the average blade area of the fully expanded fifth leaves compared with wild-type leaves (Fig. 7a,b), suggesting that OSR1 and ARGOS are functionally redundant. As a mutant for the ARL gene is not available, we generated RNA-interfered ARL transgenic plants (ARLi) in the osr1-1 argos-1 background, and found that underexpression of ARL resulted in just a slight further reduction in the average size of osr1-1 argos-1 leaves (Fig. 7a,b, S4c). Nonetheless, reductions in palisade cell number and cell size were consistently observed in osr1-1 argos-1 and ARLi/osr1-1 argos-1 plants (Fig. 7b). In addition, we also overexpressed ARGOS or ARL in 35S-OSR1 plants, and found that overexpression of neither ARGOS nor ARL further increased the size of 35S-OSR1 organs (Fig. S5a,b). Meanwhile, introduction of the osr1-1 mutation into 35S-ARGOS plants or the argos-1 mutation into 35S-OSR1 plants did not attenuate organ enlargement in 35S-ARGOS or 35S-OSR1 plants (Fig. S5c). These observations further support the hypothesis that these three genes are functionally redundant in organogenesis.
OSR1 is responsive to plant hormones and co-regulated with ARGOS and ARL
As ARGOS is induced by auxin and cytokinin and ARL by brassinosteroid (Hu et al., 2003, 2006), we investigated whether OSR1 is also responsive to plant hormones. qRT-PCR analysis in seedlings treated with various hormones revealed that OSR1 is induced by ethylene but repressed by ABA and epi-brassinolide (epi-BL) treatments (Fig. 8a), indicating that OSR1 is also regulated by plant hormones, but the responses are different from those of ARGOS or ARL.
To further elucidate the regulation of OSR genes by plant hormones, we also identified the typical hormone-responsive elements in the OSR promoter regions, using the Plant Cis-acting Regulatory DNA Elements database (PLACE) (Higo et al., 1999). As shown in Fig. 8(b), two ethylene-responsive elements of tomato E4 (EREs), two ABA-responsive element-like (ABRE-like) elements and one GA-responsive element (GARE) were found in the OSR1 promoter region. Within the ARGOS promoter, there were EREs, ABRE-like elements and GAREs as well as two typical auxin-responsive elements (AuxREs), which may confer the capacity for the induction of ARGOS by auxin, while a GARE and an ABRE-like element were identified in the ARL promoter region.
Because OSR1, ARGOS and ARL regulate organ growth in a redundant manner, we investigated whether they interact and whether their transcription is coordinated during organogenesis. We could not detect any in vitro interactions among them using the yeast two-hybrid assay (Fig. S6), suggesting that they may not function through homo- or heterodimerization. Interestingly, although we could not detect an obvious alteration of ARGOS and ARL mRNA levels in the osr1-1 mutant, overexpression of OSR1 resulted in suppression of ARGOS and ARL transcription (Fig. 8c). Similarly, overexpression of ARGOS also inhibited ARL expression (Fig. 8c). This finding suggests that the three OSR genes are co-regulated during organ development.
Phylogeny of OSR-like genes
Our finding suggests that the Arabidopsis OSR genes belong to a small family that functions in coordination of growth signals and development. To expand our knowledge of the evolutionary relationships among OSR genes, we performed a BLAST search with OSR domain in the databases and identified 39 predicted OSR homologues in 14 plant species. Phylogenetic analysis of these OSR homologues revealed that a single-copy OSR gene exists in the Physcomitrella patens genome, and multiple duplication events have occurred during the evolution of plant species (Fig. S7). The four Arabidopsis OSR homologues were probably generated through three gene duplication events. One gene duplication event happened before the divergence of core eudicots, giving rise to ARGOS/ARL and OSR1/At2g41225. Two recent independent gene duplications produced ARGOS and ARL, as well as OSR1 and At2g41225 (Fig. S7). At2g41225 is located alongside OSR1 in the genome, and appears to encode a protein containing an incomplete OSR domain (data not shown). In addition, although multiple OSR homologues were identified in grass species, such as Oryza sativa, Sorghum bicolor and Zea mays, they obviously experienced different evolutionary histories from Arabidopsis OSR genes (Fig. S7).
A number of genes involved in organ size control in plants have been identified, and these genes are referred to as ‘intrinsic yield genes’ (IYGs) because of their potential to boost the biomass and yield of agronomic plants (Gonzalez et al., 2009; Krizek, 2009). However, most of these genes appear to regulate organ growth through independent pathways (Busov et al., 2008; Gonzalez et al., 2009), suggesting that the molecular mechanisms of organ size control in plants are more complicated than those in animals. Here, we identified the OSR1 gene, which regulates organ growth by affecting both cell proliferation and cell expansion, and functions through prolonged expression of ANT, as does ARGOS. OSR1 shares low amino acid homology with ARGOS and ARL, but they do contain a relatively conserved OSR domain that is sufficient to promote organ growth. Although the molecular and biochemical mechanisms are still unclear, our characterization of OSR1, together with the previous work on ARGOS and ARL, reveals a small family of proteins that influence organ growth and final organ size by regulating cell proliferation and/or cell expansion.
In plants, organ growth is largely influenced by intrinsic and environmental signals, including plant hormones. Most plant hormones, such as auxin, cytokinin, BR and ethylene, have been demonstrated to play an important role in regulating cell proliferation and/or cell expansion (Tsukaya, 2002b, 2006). However, how these hormonal signals are integrated into the developmental program remains largely unknown. Although our observations indicate that the three OSR genes are functionally redundant, their transcription is regulated by different plant hormones. ARGOS is strongly induced by auxin and cytokinin, and the induction depends on AUXIN-RESISTANT1 (AXR1) (Hu et al., 2003). ARL, in contrast, is induced by BR, and this induction relies on the BR receptor BRI1 (Hu et al., 2006). Furthermore, OSR1 is up-regulated by ethylene but repressed by ABA and BR. These findings suggest that these plant hormones regulate organogenesis by modulating the expression of different OSR genes that perform similar molecular functions.
During organogenesis, growth of an organ to its characteristic size involves two overlapping and coordinated processes; that is, cell proliferation and cell expansion. For example, cell expansion and cell proliferation occur concurrently in a growing Arabidopsis leaf (Donnelly et al., 1999). However, cell division and cell expansion are generally considered to be separately controlled (Neufeld et al., 1998). Indeed, most identified factors involved in plant organ size control have been found to affect either cell proliferation or cell expansion; only a few factors, such as ARF2, ERBB-3 Binding Protein 1 (EBP1) and HERCULES1 (HRC1), have an impact on both cell proliferation and cell expansion (Horváth et al., 2006; Century et al., 2008; Gonzalez et al., 2009). Most importantly, a compensatory mechanism between cell expansion and cell proliferation has often been found in studies of plant organ development (Tsukaya, 2002a). Although the OSR genes function in a redundant way at the organ level, the cellular mechanisms underlying such functions are apparently different among the three genes. ARGOS mainly affects the duration of cell proliferation (Hu et al., 2003), whereas ARL mainly influences cell expansion (Hu et al., 2006). OSR1 predominantly regulates cell proliferation but also affects expansion. In addition, overexpression of OSR1 or ARGOS suppresses the expression of ARGOS and/or ARL, suggesting that these genes are co-regulated in plants. Therefore, our results, together with the recent work on OsARGOS (Wang et al., 2009), suggest that the members of the OSR family may be co-evolved factors that are involved in the coordination of cell proliferation and expansion in plant development.
Taken together, our results define a plant-specific pathway by which the co-evolved members of the OSR family integrate plant growth signals into the regulation of cell proliferation and expansion during organ growth (Fig. S8). This pathway, presumably, acts together with other identified or as yet unknown pathways, to control organ development and determine overall organ size in plants. However, the molecular mechanism by which OSR genes regulate cell proliferation and/or cell expansion remains unclear. OSR proteins are ER-localized, and they must function together with other molecules to regulate the expression of ANT and/or other genes that affect cell division and expansion machineries. Thus, identification of the proteins interacting with OSR will further our understanding of how cell proliferation and cell expansion are coordinated in plant growth and development. Finally, the increase in biomass in OSR1-overexpressing plants suggests that OSR1 or its orthologues may be potential targets for genetic engineering to increase crop yields.
We are grateful to Dr Ming Yang (Oklahoma State University) for critical reading of the manuscript. We thank Drs Robert L. Fischer for kindly providing ant-1 seeds, and Jing Bo Jin and Nam-Hai Chua for vectors used in this study. We thank Drs Jing Bo Jin and Guosheng Xiong for their technical assistance with protoplast transformation and confocal microscopy, and Zheng Meng and Xuelei Lin for their assistance with, and discussions concerning, phylogenetic analysis. This work was supported by the Ministry of Science and Technology of China (2009CB941500 and 2007CB948200) and the National Natural Science Foundation of China (30821007), and a grant from the CAS/SAFEA International Partnership Program for Creative Research Teams.