OsORC3 is required for lateral root development in rice


For correspondence (e-mail mcz@zju.edu.cn).


The origin recognition complex (ORC) is a pivotal element in DNA replication, heterochromatin assembly, checkpoint regulation and chromosome assembly. Although the functions of the ORC have been determined in yeast and model animals, they remain largely unknown in the plant kingdom. In this study, Oryza sativa Origin Recognition Complex subunit 3 (OsORC3) was cloned using map-based cloning procedures, and functionally characterized using a rice (Oryza sativa) orc3 mutant. The mutant showed a temperature-dependent defect in lateral root (LR) development. Map-based cloning showed that a G→A mutation in the 9th exon of OsORC3 was responsible for the mutant phenotype. OsORC3 was strongly expressed in regions of active cell proliferation, including the primary root tip, stem base, lateral root primordium, emerged lateral root primordium, lateral root tip, young shoot, anther and ovary. OsORC3 knockdown plants lacked lateral roots and had a dwarf phenotype. The root meristematic zone of ORC3 knockdown plants exhibited increased cell death and reduced vital activity compared to the wild-type. CYCB1;1::GUS activity and methylene blue staining showed that lateral root primordia initiated normally in the orc3 mutant, but stopped growing before formation of the stele and ground tissue. Our results indicate that OsORC3 plays a crucial role in the emergence of lateral root primordia.


The root system facilitates nutrient and water uptake and is consequently important for plant growth and reproduction. The lateral root (LR) is a main component of both the tap root and fibrous root systems. LR formation is a complex developmental process that is regulated by both hormones and environmental signals. A series of well-characterized and highly coordinated cell divisions give rise to formation of LR primordia (LRPs) (Malamy and Benfey, 1997; Peret et al., 2009). After initiation, the LRP emerges and breaks through the epidermal cells to form the LR (Malamy and Benfey, 1997; Nibau et al., 2008). Whereas LR development has been studied extensively in the dicot model plant Arabidopsis (Peret et al., 2009; Benkova and Bielach, 2010), our understanding of this process is limited in monocot plants, including the model cereal plant rice (Oryza sativa). Although several lateral rootless mutants have been reported in rice (Rebouillat et al., 2009), only three genes have been functionally identified as controlling LR development in this species, and all three encode Aux/IAA proteins (Nakamura et al., 2006; Song et al., 2009; Zhu et al., 2011). The molecular mechanisms underlying LR initiation and development in monocot plants are largely unknown.

The origin recognition complex (ORC) consists of six proteins (ORC1–ORC6) in yeast and humans, and plays a fundamental role in DNA replication in all eukaryotes (Bell, 2002). The ORC binds directly to replication origins to form pre-replication complexes that recruit cell division cycle 6 (CDC6), chromatin licensing and DNA replication factor 1 (CDT1) and mini-chromosome maintenance proteins 2–7 (MCM2-7) proteins upon exit from mitosis. The ORC is generally conserved in fungi, plants and animals (DePamphilis et al., 2006; Sasaki and Gilbert, 2007). Recent research has shown that mutations in the ORC1, ORC4 and ORC6 subunits cause human Meier–Gorlin syndrome, which is characterized by primordial dwarfism and other developmental abnormalities. It has been suggested that a defect in DNA replication, rather than in the DNA replication-independent functions of ORC, is responsible for Meier–Gorlin syndrome (Sasaki and Gilbert, 2007).

In addition to their roles in DNA replication, ORC proteins have been shown to be involved in processes such as the establishment of mitotic chromosome structure (Loupart et al., 2000; Prasanth et al., 2002, 2004; Sanchez and Gutierrez, 2009), regulation of transcription (Pak and Sheng, 2003; Prasanth et al., 2004), sister chromatin cohesion, cytokinesis and neural dendritic branching (Rusche et al., 2003; Sasaki and Gilbert, 2007). In non-proliferating cerebellar granules, ORC3 supports neuronal maturation by inhibiting the Rho signaling pathway, and mediates the differentiating activity of type 4 metabotropic glutamate (mGlu4) receptors in cultured cerebellar granule cells (Cappuccio et al., 2010).

Typically, mutations in ORC genes arrest the cell cycle at the G1/S transition. However, in yeast and Drosophila melanogaster, many cells progress through S phase to mitosis (M phase), where they arrest and die (Bell et al., 1993; Loupart et al., 2000). Although much is known about ORC genes in yeast and animal model systems, little is known about plant ORC components. Homologs of all six ORC subunits from Arabidopsis and rice have been described (Kimura et al., 2000; Diaz-Trivino et al., 2005; Li et al., 2005; Mori et al., 2005; Shultz et al., 2007), and maize (Zea mays) cDNAs of ORC1–5 have been cloned (Witmer et al., 2003). In Arabidopsis, only two ORC genes (AtORC1 and AtORC2) have been functionally described (Collinge et al., 2004; Sanchez and Gutierrez, 2009). AtORC1 is reported to regulate transcription mediated by plant homeodomain (PHD) motif, and depends on ORC1 binding at target promoters. AtORC2 knockout causes a zygotic lethal phenotype in which mutant embryos and endosperm abort at an early developmental stage (Collinge et al., 2004). In rice, OsORC2 is strongly expressed in roots, seedlings and the inflorescence meristem, but its expression level is much lower in mature leaves and shoots (Li et al., 2005). OsORC1 transcripts are abundant in root tips, scarce in young leaves, and absent in mature leaves (Kimura et al., 2000). Apart from these findings, the function of ORCs has not been well characterized in plants. Furthermore, no ORC member has been functionally described in rice.

In this study, we isolated a rice mutant with defects in LR formation from an ethyl methanesulfonate-generated mutant library of the indica rice cultivar Kasalath. The mutant exhibited a temperature-dependent LR development defect. At 34°C, LRs failed to develop in the orc3 mutant, but LRPs initiated normally. However, LR development was normal at 26°C. Map-based cloning and complementation experiments indicated that the lateral rootless mutant phenotype (at 34°C) was caused by a mutation in ORC3. Knockdown of OsORC3 by RNA interference (RNAi) showed a dwarf growth phenotype and defects in LR development that were independent of temperature. Cell vitality was reduced and cell death was severely enhanced in the root tips of ORC3 RNAi rice. These results indicate that OsORC3 plays a crucial role in LR development.


The orc3 mutant exhibits a defect in LR development

A rice mutant defective in LR formation was isolated from an ethyl methanesulfonate-mutated population of the indica cultivar Kasalath. The mutant was named orc3 based on the identity of the cloned gene, OsORC3 (see below). When grown in hydroponic solution for 7 days at 34°C, the orc3 mutant showed normal primary root growth, but, in contrast to the wild-type (WT), lacked LRs (Figure 1a–e). Interestingly, LR development in the orc3 mutant was sensitive to temperature. Below 26°C, plant growth and LR development were similar in the mutant and WT (Figure 1g–j); however, no LRs were visible at temperatures above 34°C, and LR number was reduced with increasing temperature when the temperature was increased from 26 to 34°C (Figure S1). Additionally, orc3 showed a significant reduction in plant stature and seed setting when grown on soil at 34°C/26°C (12 h day/12 h night) (Figure 1f,j–l). However, the anthers and stigma developed normally, and pollen viability was not significantly different in the orc3 mutant compared with the WT at both 34°C/26°C and 30°C/22°C (Figure S2).

Figure 1.

Phenotypic characterization of the orc3 mutant. (a–e, g–i) Root phenotype of 7-day-old Kasalath wild-type (WT) plants (a–c, g, h) and orc3 mutant plants (a, d, e, g, i) grown in hydroponic solution at 34°C (a–e) and 26°C (g–i). (b–e, h, i) LRPs, as revealed by methylene blue staining, in 5-day-old WT seedlings (b, c, h) and orc3 seedlings (d, e, i) grown at 34°C (b–e) and 26°C (h, i). The root tips (b, d) include the segment from 0 to 1 cm away from the tip, whereas the root segments (c, e, h, i) are the segments from 2 to 4 cm away from the tip. Scale bars = 2 cm (a, g) and 0.5 mm (b–e, h, i). (f, j) The phenotype of WT and orc3 plants grown in soil at 34°C/26°C (day/night) (f) and 26°C/26°C (day/night) (j) for 2 months. Scale bar = 5 cm. (k) Panicle and seeds of the WT and orc3 mutant grown at 34°C/26°C. Top scale bar = 2 cm (panicle). Bottom scale bar = 0.5 cm (seed). (l) Seed setting rate of the WT and orc3 grown at 34°C/26°C. (m) Number of LRPs in the primary root (0–3.5 cm from the root tip) of 7-day-old WT and orc3 seedlings grown at 34°C/26°C. In (l) and (m), error bars represent the standard deviation (SD) from 20 roots.

We then investigated whether LRPs were initiated on the primary roots of the orc3 mutant at 34°C, using methylene blue staining. The staining pattern of the primary root meristem, LRP distribution and LRP number were similar in 5-day-old seedlings of the orc3 mutant (0–3.5 cm from the tip) and WT (Figure 1b–e,h–i,m). In the maturation zone, staining was visible in the LR tips of the WT, but no LRs or staining were observed in the orc3 mutant (Figure 1c,e).

Map-based cloning and characterization of OsORC3

To clone the mutated gene that results in the orc3 mutant phenotype, an F2 population was developed by crossing the mutant with the japonica rice cultivar Nipponbare. Segregation of WT and LR-defective plants among the resulting 240 F2 progeny showed a ratio of 3:1 (174 WT plants and 66 LR-defective plants, χ2 = 0.8 < inline image = 3.84; < 0.05), indicating that the LR defect in the orc3 mutant is caused by a single recessive gene. The gene locus was initially mapped to the long arm of chromosome 10, between SSR markers RM3311 and RM271, using 30 F2 mutant plants. Fine mapping using 2680 F2 mutants further delimited the gene to a 45 kb region between STS1 and STS4 in BAC clone OSJNBa0044A10 (Figure 2a). Based on the annotation in the Institute for Genomic Research (TIGR) database, all six putative open reading frames (ORFs) identified in this region in the orc3 mutant were sequenced. Sequence analysis revealed a nucleotide substitution in a putative gene, Os10g26280, which was annotated as OsORC3. The G nucleotide at position 3242 bp downstream of the ATG start codon, which is in the 9th exon of OsORC3, is substituted by an A residue, resulting in substitution of Cys (C) by Tyr (Y) at amino acid 432 (Figure 2b). The G→A substitution generates an ScaI enzyme recognition site in the orc3 mutant, which was used to identify the orc3 mutant using a CAPS marker (Figure 2d).

Figure 2.

Map-based cloning and complementation analysis. (a) Map-based cloning of ORC3. ORC3 was first mapped between markers RM3311 and RM271 on chromosome 10, and finally between STS1 and STS4. Numbers below the molecular marker indicate the number of recombinant(s) in the orc3/Kas F2 population. The point mutation in the 9th exon is indicated. The gene model of ORC3 is shown at the bottom. Black boxes indicate exons, black lines indicate introns and the 5' and 3' untranslated regions. (b) Secondary structure analysis of OsORC3 using the PROSITE database (http://prosite.expasy.org/). The point mutation resulting in a substitution of cysteine (C) with tyrosine (Y) at amino acid 432 (C432Y) is indicated. The ORC3_N domain is indicated at the N-terminus. (c) Root phenotype of the wild-type (WT), orc3 mutant and two transgenic ORC3 complementation lines (in an orc3 mutant background; designated OX1 and OX2). Scale bar = 1 cm. (d) CAPS marker analysis of the transgenic lines. The 705 bp band represents the undigested PCR amplicon. (e) ORC3 transcript level in the transgenic lines, as revealed by RT-PCR analysis. OsACTIN1 was used as the internal control. (f) Analysis of LR number in the WT, orc3 mutant and OsORC3 complementation lines. Error bars indicate the standard deviation (SD) from 30 roots.

The OsORC3 ORF is 2061 bp and encodes a 686 amino acid protein. A typical ORC3-N domain is located at the N-terminus (amino acids 46–365), as revealed by domain analysis using the PROSITE database (http://prosite.expasy.org/; Figure 2b). Phylogenetic analysis using other known ORC3 subunits showed that OsORC3 is more similar to monocot proteins such as maize ZmORC3 (66.85% amino acid identity) and sorghum (Sorghum bicolor) SbORC3 (70% identity) than to dicot proteins such as Arabidopsis (41.2% identity) (Figure S3a). The mutated amino acid C is conserved in plants and humans (Figure S3b), suggesting a conserved function of this amino acid in the plant and animal kingdom.

To confirm the mapping result, we performed a complementation test in which full-length WT OsORC3 cDNA driven by the 35S promoter was expressed in the orc3 mutant. More than ten transgenic lines were identified by RT-PCR and Southern blot analysis, all of which showed normal LR development, even at 34°C, indicating complete rescue of the mutant phenotype (Figure 2c). Two representative independent transgenic lines (designated OX1 and OX2) are shown in Figure 2(c–f) and Figure S4. Moreover, the growth performance and number of LRs in the complementation lines were indistinguishable from those of the WT (Figure 2c–f). These results confirm that the mutant phenotype is caused by mutation of OsORC3.

Abnormal mitotic activity and LRP development in the orc3 mutant

CYCB1;1 is strongly expressed in cells with high mitotic activity (Doerner et al., 1996), and has been used as a marker of G2/M of the cell cycle (Colón-Carmona et al., 1999). To investigate whether the abnormal LRPs in the orc3 mutant were caused by abnormal mitotic activity, the orc3 mutant was transformed with CYCB1;1::GUS, and GUS activity was further examined. As shown in Figure 3, GUS staining resulted in distinctive blue dots in the LRPs and LR tips of WT seedlings, where cell division occurred at a high frequency (Figure 3a–c). Although patches of GUS staining were observed in orc3 mutants that lacked LRs, staining was smeared around the LRP zone (Figure 3d–f). Staining was absent in the mature zone of orc3 primary roots (Figure 3f). LRP development is divided into eight stages in Arabidopsis. At stage IV, the LRP forms four layers by periclinal divisions of the inner layer. LRPs are midway through the parent cortex by stage V, and they finally emerge at stage VIII (Casimiro et al., 2003). Longitudinal sections showed that LRP development in the orc3 mutant was normally initiated and remained normal until stage IV, after which the LRPs degenerated (Figure 3l–p).

Figure 3.

LRP development and expression analysis of cell cycle-related genes. The CYCB1;1::GUS construct was transformed into the wild-type (WT) and orc3 mutant by A. tumefaciens-mediated transformation. (a–f) LRPs in 5-day-old WT seedlings (a–c) and orc3 mutant seedlings (d–f) grown in hydroponic solution. (g–p) LRP initiation and formation were examined in longitudinal sections through the LRP formation zone (2–3 cm from the root tip) of 5-day-old Kasalath (WT) seedlings (g–k) and orc3 seedlings (l–p), respectively. Scale bars = 0.4 mm (a–f). (q) Cell cycle-related gene expression, as analyzed by quantitative RT-PCR. Error bars indicate the standard deviation (SD) from three biological repeats. Asterisks indicate a statistically significant difference between the WT and orc3 at < 0.05 (Student's t test).

Cyclins and cyclin-dependent kinases (CDKs) are key regulators in determining a cell's progression through the cell cycle (Nigg, 1995). In yeast, the ORC is required at late G1 for the G1/S transition and formation of the pre-replication complex (Makise et al., 2008). Therefore, expression of cell cycle-related genes in WT and orc3 rice roots was further investigated to explore the possible reasons for aberrant LR development. At 26°C, the expression level of the examined genes was similar in the WT and the orc3 mutant (Figures S5 and S6a), in agreement with their similar phenotype, including LR development. At 34°C, however, the expression level of the cyclins CYCD2;1, CYCD2;2 and CDKB1 was significantly reduced in the orc3 mutant, while that of CYCB1;1 and KRP1 was elevated (Figure 3q). The relative expression of the cyclin genes at 26 and 34°C was also examined. In the WT, expression of CYCB1;1 was lower at 34°C than at 26°C, whereas expression of CYCD2:1, CYCD2:1, CDKB1 and KRP1 was slightly higher at 34°C than at 26°C (Figure S6b). Given the defect in LRP development and the pattern of cyclin expression in the orc3 mutant, it is reasonable to speculate that mutation of ORC3 disrupts cell-cycle progression and then blocks development of LRs before they extend from the epidermal cells.

Expression pattern and subcellular localization of OsORC3

Real-time quantitative RT-PCR analysis showed that OsORC3 was highly expressed in the root, young shoot, stem base and young panicle, but weakly expressed in the mature leaf (Figure 4a). Tissue-specific expression was investigated by generating transgenic plants that harbored the OsORC3p:GUS vector, in which the uidA reporter gene is fused to the OsORC3 promoter. Ten independent ORC3p:GUS transgenic lines were analyzed for GUS activity. OsORC3 was expressed in the primary root tip, LRP, emerged LR tip, stem base, young leaf, anther and ovary (Figure 4b–i), where cells are highly proliferative. This result is consistent with a previous report showing that OsORC3 is expressed in tissues undergoing active cell proliferation (Mori et al., 2005). In situ hybridization analysis further confirmed that OsORC3 was expressed in the LRP, young leaf and stem base, where adventitious root primordia initiate (Figure 4j–o). The specific expression of OsORC3 in the root tip and LRPs suggests that ORC3 is involved in the development of root systems.

Figure 4.

Spatial expression pattern and subcellular localization of OsORC3. (a) OsORC3 expression pattern, as determined by quantitative RT-PCR. Total RNA was extracted from the root, stem base, young shoot, mature leaf and panicle of the wild-type. OsACTIN1 was used as the internal control. The expression level in the young shoot was set at 1. Values are means and standard deviation (SD) of three biological repeats. (b–i) The expression pattern of OsORC3, as revealed by promoter–GUS fusion analysis in transgenic seedlings. GUS activity was observed in the tip of the primary root (b), LRPs (c), emerging LR (d), the tip of fully emerged LR (e), stem base (f), young leaf (g), young spikelet (h) and anthers and ovary (i). Scale bars = 0.2 mm (b, c) and 1 mm (d–i). (j–o) OsORC3 expression pattern, as revealed by in situ hybridization. (j–l) Hybridization using the antisense probe of the stem base (j), LRP (k) and young shoot (l). Arrows in (j) and (m) indicate the adventitious root primordia. (m–o) Hybridization using the sense probe of the stem base (m), LRP (n) and young shoot (o). Scale bars = 200 μm (j, m) and 50 μm (k–l, n, o). (p–u) Transient expression of the ORC3–GFP fusion (s–u) and GFP only (p–r) in onion epidermal cells. Full-length OsORC3 cDNA was fused to the N-terminus of GFP and transiently expressed in the onion epidermal cells for 12–16 h. The photographs were taken under a dark field by fluorescence microscopy (p, s) and under a bright field to examine the morphology of the cell (q, t). Merged images are shown in (r) and (u).

The subcellular localization of OsORC3 was determined by transiently expressing OsORC3:GFP under the control of the CaMV 35S promoter in onion epidermal cells. The green fluorescent protein (GFP) signal was exclusively detected in the nuclei, indicating that OsORC3 is located in the nucleus (Figure 4p–r), as are OsORC1 and OsORC2 (Mori et al., 2005), in accordance with reports that ORC functions in DNA replication within the nucleus (Sasaki and Gilbert, 2007).

ORC3 knockdown transgenic rice showed retarded growth and LR defects

To further investigate the physiological and molecular function of OsORC3, we developed RNAi transgenic plants with reduced expression of OsORC3 by introducing artificial microRNA. More than 20 transgenic lines were investigated. RNAi lines showed striking growth phenotypes, including LR defects, severely reduced stature and short roots. Two representative lines, RNAi#1 and RNAi#2, are shown in Figure 5, and were used for the RNAi study. Expression of OsORC3 was examined in ORC3 knockdown transgenic lines by RT-PCR analysis. In the RNAi lines, the level of OsORC3 transcript was significantly reduced compared with the WT (Figure 5k).

Figure 5.

Knockdown of OsORC3 causes a lateral rootless and dwarf phenotype. (a–g) Phenotype of 1-week-old wild-type (WT) and RNAi lines (RNAi#1 and RNAi#2) grown in hydroponic solution at 30°C/24°C. (a, e) Base part of the root in WT (a) and RNAi (e) seedlings. (b, c, f, g) LRPs of WT (b, c) and RNAi (f, g) lines were investigated by methylene blue staining. (h, i) Transverse sections through the primary root, showing LRPs (indicated with red arrows) in the WT (h) and ORC3 RNAi (i) lines. The dashed rectangles in (b) and (f) show the regions examined in these transverse sections. (j) Comparison of shoot height and primary root length in the WT and RNAi lines. Error bars indicate the standard deviation (SD) of 15 independent seedlings. (k) OsORC3 expression level, as revealed by quantitative RT-PCR, in WT and ORC3 RNAi seedlings. Total RNA was extracted from the roots of 7-day-old seedlings. (l) Number of LRPs in the WT and ORC3 RNAi seedlings. Error bars indicate the standard deviation (SD) of 15 independent seedlings. Asterisks in (j) and (l) indicate a statistically significant difference compared with WT at < 0.01.

After 7 days of hydroponic growth, the primary roots and shoots of the ORC3 RNAi plants were only approximately 50% as long as those of the WT (Figure 5d,j). Furthermore, in contrast to the orc3 mutant, the LR defect in ORC3 knockdown rice was not sensitive to growth temperature. Almost no LRPs were observed in 1-week-old ORC3 RNAi seedlings using methylene blue staining (Figure 5l). However, tiny LRPs were observed in cross-sections of the RNAi lines (Figure 5i). Rarely, a few LRs were found in those lines in which OsORC3 expression was not strongly knocked down.

OsORC3 RNAi rice seedlings exhibit reduced cell division activity

Cell division and cell expansion play key roles in the final stature of the plant. Given the significantly dwarfed stature of ORC3 knockdown rice and the known function of the ORC complex, we examined the number and size of cells in the root tip of WT and ORC3 RNAi seedlings to understand whether the retarded growth was due to a decrease in the number or size of cells (Figure 6). Inspection of confocal images of longitudinal sections through the primary root revealed that the size of the meristematic zone in ORC3 RNAi roots was smaller than that of the WT (region between the long and the short arrows in Figure 6a,b). In addition, the cells in the meristematic zone of the ORC3 RNAi lines were larger than those in the WT (Figure 6a,b), and thus the number of cells in this region of the ORC3 RNAi line was approximately 30% less than in the WT (Figure 6i). Furthermore, cells in the root apical meristem of the ORC3 RNAi line were not uniformly organized. All these results suggest that cell division was disturbed by knockdown of ORC3. On the other hand, the length of cells in the maturation zone was much shorter (approximately 50%) in the ORC3 RNAi line than in the WT (Figure 6j), indicating that cell elongation was also affected in the ORC3 RNAi root.

Figure 6.

Cell structure and reduced vital activity in the root tips of OsORC3 knockdown rice. (a, b) Propidium iodide-stained primary root tips of 5-day-old WT (a) and ORC3 RNAi (b) lines grown at 30°C/24°C (day/night). Short arrows indicate the start position of the elongation zone of the primary root. Long arrows indicate quiescent center cells. Scale bars = 50 μm. (c–f) Evans blue staining of the root tip and maturation zone of the WT (c, d) and orc3 (e, f). Scale bars = 1 mm. (g, h) Feulgen staining of 5-day-old WT (g) and OsORC3 RNAi (h) roots. Scale bar = 200 μm. (i) Cell number in the meristem area. Five-day-old primary roots were used to analyze cell number. The number of cells in the meristem zone [the region between the long and short arrows in (a) and (b)] of 10 primary roots was counted in a 50 μm × 50 μm area using confocal microscopy images of propidium iodide-stained roots. Eight areas were analyzed per root. Error bars indicate standard deviation (SD). (j) Cell length in the maturation zone of WT and OsORC3 RNAi roots. The cell length of cortical cells in the maturation zone of 15 seedlings was assessed using confocal microscopy images of propidium iodide-stained primary roots of 5-day-old seedlings. Asterisks in (i) and (j) indicate a statistically significant difference compared with WT at < 0.01.

We further examined root cell proliferation activity in vivo. Evans blue staining, which detects vital activity, revealed a large zone in the root tip of ORC3 RNAi lines, but not in the WT, where cell mitotic activity is high to maintain rapid cell division (Figure 6c–f). This observation indicates that knockdown of OsORC3 results in cell death in the root meristematic zone. Staining with Feulgen stain, which selectively stains DNA, was further used as an indicator of DNA content (Demchenko et al., 2004). A red region in the meristematic zone was observed in WT root tips after Feulgen staining (Figure 6g), but not in ORC3 RNAi root tips (Figure 6h). These results suggest that the cell proliferation activity was lower in the root tips of ORC3 RNAi lines than in those of the WT.


The hexameric ORC, a well-known component of eukaryotic DNA replication, is conserved in all eukaryotes including rice (Bell, 2002; Mori et al., 2005). Although much is known about the ORC subunits in yeast and animal model systems, understanding of the function of ORC subunits in plants is still very limited.

OsORC3 is required for LR development in rice

The de novo organogenesis of LRs gives rise to the ideal root architecture for water and nutrient acquisition in plants. Generally, LR development is divided into two stages: LRP initiation and LR emergence (Bhalerao et al., 2002). However, it may also be divided into eight stages (Casimiro et al., 2003). In rice, LRPs are initiated by anticlinal cell divisions, followed by periclinal cell divisions in the pericycle and endodermis (Kawata and Shibayama, 1965). In the orc3 mutant, no visible LRs emerged above 34°C; however, a similar number of LRs as the WT emerged at 26°C or lower. Methylene blue staining indicated that the number and distribution of LRPs in the primary roots of the orc3 mutant (0–3 cm from the tip) was similar to that of the WT in 5-day-old seedlings at 26°C (Figure 1b,d,m). RNAi transgenic lines in which ORC3 was knocked down lacked visible LRs and had a severely reduced stature in both the shoot and root. Almost no LRPs were observed in 1-week-old ORC3 RNAi lines subjected to methylene blue staining. However, very early-stage LRPs were observed in cross-sections of roots analyzed by microscopy (Figure 5i). These results indicate that LRP emergence is blocked in orc3 and RNAi lines, and that LRPs stopped growing at an earlier stage in the RNAi lines than in the orc3 mutant. Rarely, a few LRs were observed in knockdown lines that exhibited moderate OsORC3 expression, suggesting that a greater loss of function correlates with a more severe phenotype. These results indicate that OsORC3 is necessary for LRP development and LR emergence.

In the orc3 mutant, LRPs initiate regularly but stop growing in a temperature-dependent manner (Figure 1). Expression analysis indicated that ORC3 transcript levels did not change at different growth temperatures (Figure S5). Furthermore, in contrast to the orc3 mutant, the LR defect in ORC3 knockdown rice was not sensitive to temperature. The orc3 mutant was caused by a point mutation: substitution of cysteine for tyrosine at amino acid 432 of OsORC3. The DiANNA (http://clavius.bc.edu/~clotelab/DiANNA/) and DIpro2 (http://download.igb.uci.edu/intro.html) prediction software predict that Cys432 forms a disulfide bond with two Cys residues in ORC3 (i.e. Cys432–Cys493 and Cys432–Cys449). The Cys432 to Tyr432 mutation breaks down these two disulfide bonds in the orc3 mutant, which may cause a mild change in the three-dimensional structure of the ORC3 protein or the pre-replication complex at high temperatures (i.e. 34°C and above). This change may eventually affect the function of ORC3, making the orc3 mutant temperature-sensitive. In Arabidopsis, a T-DNA mutant disrupted in the expression of AtORC2 exhibited a zygotic lethal phenotype in which embryos aborted at the eight-cell stage, which hampered a detailed functional investigation of AtORC2 (Collinge et al., 2004). In this study, we also found that severely knocked down RNAi lines did not survive. The orc3 mutant, which has a temperature-sensitive defect in LR development, thus represents a powerful tool for further investigating the function of ORC3 in vivo.

In Drosophila, mutation of ORC3 reduces the overall volume of the mushroom body, a specific anatomical region of the adult brain that is thought to be involved in olfactory learning. The proposed reason for this is that the mutation results in a rate of cell division that is not uniform when the brain requires a faster cell division cycle (Pinto et al., 1999). Mutation of OsORC3 caused a defect in LR development, but not of other tissues at the early growth stages. Cells of the LRP differentiate almost immediately after initiation, and many highly proliferative cells are required for proper cell division and expansion of this tissue (Malamy and Benfey, 1997). Therefore, the finding that cell division does not proceed at a uniform rate in the orc3 mutant may also account for the defect in LR development, which requires rapid cycles of cell division.

ORC3 affects cell division and cell viability in rice

The meristematic zone of roots of ORC3 RNAi plants was smaller than that of the WT (Figure 6a,b) and had approximately 30% fewer cells than the corresponding region of the WT (Figure 6i). Furthermore, cells in the root apical meristem of the ORC3 RNAi line were not uniformly organized (Figure 6b). This finding suggests that cell division is disturbed in the RNAi lines. At 34°C, LRP development was blocked in the orc3 mutant before the stele and ground tissues formed (Figure 3l–p). Furthermore, the expression level of the cyclins CYCD2;1, CYCD2;2 and CDKB1 was significantly reduced in the orc3 mutant at 34°C, whereas CYCB1;1 and KRP1 accumulated (Figure 3q). The D-type cyclins, CYCDs, have been reported to control entry into the mitotic cell cycle and the transition of cells from G1 phase to S phase through their association with cyclin-dependent kinases A (Riou-Khamlichi et al., 2000). Transgenic expression of Arabidopsis CYCD2 in tobacco roots shortens the G1 phase (Cockcroft et al., 2000). KRP1, which inhibits the cell cycle and regulates the G1/S transition, is strongly down-regulated during the G1 phase in cell cycle re-entry (Ren et al., 2008). KRP1 over-expression was shown to inhibit cell division and LR formation (Wang et al., 2000). The CYCD2/CDKA complexes sequester ectopically expressed KRP1 (Ren et al., 2008). Thus, the up-regulation of KRP1 and down-regulation of CYCD2 in the orc3 mutant grown at 34°C suggest that cell division was arrested at the G1/S transition. This is consistent with the report that mutations in ORC genes arrest cell division at the G1/S transition of the cell cycle (Makise et al., 2008). CYCB1;1 and CDKB1 have been reported to mark the G2/M step of the cell cycle (Colón-Carmona et al., 1999; Vanneste et al., 2005). Plants with reduced CDKB1 activity prematurely exit the mitotic cell cycle (Boudolf et al., 2004). The up-regulation of CYCB1;1 and down-regulation of CDKB1 in the orc3 mutant grown at 34°C suggest that the G2/M transition may occur prematurely in the orc3 mutant at 34°C. Given the pattern of LRP development and cell-cycle gene expression in the orc3 mutant, it is reasonable to speculate that mutation of the ORC3 subunit disturbs cell-cycle progression and then blocks LR development before the lateral root extends from the epidermal cells.

On the other hand, the increased Evans blue staining in the root tip of ORC3 RNAi lines but not those of the WT indicates that knockdown of OsORC3 results in cell death in the root meristematic zone. Almost no staining with Feulgen stain, which is DNA-selective, was observed in the root tip of RNAi lines, suggesting that cell proliferation activity was lower in the root tips of the ORC3 RNAi lines than those of the WT. All these results suggest that ORC3 functions in DNA replication, cell proliferation and viability, and are in agreement with the functions of ORCs reported for yeast and human systems (Bicknell et al., 2011a,b). The orc2-1 mutation was reported to possibly trigger apoptosis in yeast by excessive checkpoint activation (Weinberger et al., 2005). In the present study, LRPs were initiated in the orc3 mutant and RNAi lines, but degenerated at a later stage, suggesting that loss of function of OsORC3 may also cause apoptosis of cells in the LRP in rice. Further studies are required to investigate why the LRP develops abnormally in response to mutation of ORC3.

Altogether, OsORC3 mutations cause a broad spectrum of phenotypes, including dwarfism, lack of LRs and short primary roots. Our results indicate a direct function for OsORC3 in LR development in rice.

Experimental procedures

Plant materials, growth conditions and mutant screen

The rice (Oryza sativa) mutant orc3 was identified at 34°C from an ethyl methanesulfonate-mutated population of the indica cultivar Kasalath grown hydroponically as described by Yoshida et al. (1976). For hydroponic growth, plants were grown in culture solution in a growth chamber at the specified temperature, a humidity of 70%, and a 12 h light/12 h dark cycle (15 000 lux), except where otherwise specified. In soil-grown experiments, plants were grown under a 34°C/26°C or 30°C/24°C (day/night) regime in a greenhouse on the Zijingang Campus, Zhejiang University (Hangzhou, China).

Map-based cloning of the OsORC3 gene

A mapping population of 2680 F2 plants was generated from crosses between the homozygous orc3 mutant and japonica variety Nipponbare. Lateral rootless plants from the F2 population were used to clone OsORC3. OsORC3 was first mapped using SSR markers and STS markers (Table S1). The ORC3 locus was finally mapped to within a 45 kb region between STS1 and STS4 in BAC clone OSJNBa0044A10 on chromosome 10. The ORF of the candidate gene was amplified from both the orc3 mutant and WT using a pair of specific primers (OsORC3 gene-F and OsORC3 gene-R) (Table S1). The PCR products were sub-cloned into the pUCm-T vector (Shanghai Sangon Biotech Co. Ltd, http://www.sangon.com/) and sequenced using an ABI 3710 sequencer (Applied Biosystems, http://www.appliedbiosystems.com.hk/).

Reverse transcriptase-PCR and quantitative RT-PCR analysis

First-strand cDNA was synthesized from total RNA using SuperScript II reverse transcriptase (Invitrogen, http://www.invitrogen.com/). RT-PCR was performed using gene-specific primers ORC3-RT-F and ORC3-RT-R. Quantitative real-time RT-PCR was performed using a Roche 480 detection system and SYBR Green I Master Mix (Roche, http://www.roche.com/) according to the manufacturer's instructions, using primers ORC3-qRT-F and ORC3-qRT-R. OsACTIN1 (LOC_Os03 g50885), amplified using primers OsACTIN1-F and OsACTIN1-R, was used as an internal control.

In situ hybridization

In situ hybridization was performed as described previously (Zhao et al., 2009), with the following modifications. The ORC3 probe was amplified using the gene-specific primers ORC3insitu-F and ORC3insitu-R (Table S1). The 760 bp PCR fragment of ORC3 cDNA was sub-cloned into the pUCm-T vector (Sangon, http://www.sangon.com/) and transcribed in vitro from either the T7 or SP6 promoter for sense or antisense strand synthesis, respectively, using a digoxigenin RNA labeling kit (Roche).

Construct design

For complementation, full-length OsORC3 cDNA was amplified using a pair of primers that included a SacI restriction site (OsORC3-F and OsORC3-R). The OsORC3 cDNA was digested by SacI after sequencing and inserted into the binary vector pCAMBIA1300 (http://www.cambia.org/) directly, such that ORC3 was under the control of a CaMV 35S promoter. The construct was transformed into rice callus developed from homozygous orc3 seeds using an Agrobacterium Tumefaciens-mediated transformation system. The transgenic plants were detected by RT-PCR.

To generate the ORC3p::GUS construct, a 2889 bp promoter region, which included the first 66 bp of the coding region of OsORC3, was amplified using a pair of specific primers: OsORC3P-F (which contains an XbaI restriction site) and OsORC3P-R (which contains a KpnI restriction site). The ORC3 promoter region was PCR-amplified, cut using XbaI and KpnI, and cloned in-frame with the uidA gene (β-glucuronidase) in pBI101.3 GUS plus (http://genomewww.stanford.edu/vectordb/vector_descrip/COMPLETE/PBI1013.SEQ.html). The resultant ORC3p::GUS construct was transformed into WT Kasalath by A. tumefaciens-mediated transformation.

The CYCB1;1::GUS fusion construct was constructed as described by Colón-Carmona et al. (1999). A translational fusion of the 2317 bp fragment upstream of the OsCYCB1;1 (LOC_Os01 g59120) start codon and a 911 bp fragment of the ORF, starting at the ATG start site, which contains a mitotic degradation box with an N-terminal 124 amino acids, was amplified using primers CYCB1;1-F and CYCB1;1-R. The 3228 bp PCR product was digested with SalI and KpnI, and inserted into the binary vector pBI101.3 GUS plus as an in-frame fusion with uidA. The CYCB1;1::GUS fusion construct was transformed into WT Kasalath and the orc3 homozygous mutant via A. tumefaciens-mediated transformation.

Subcellular localization of OsORC3

CaMV35S-ORC3-mGFP was sub-cloned into the binary vector pCAMBIA1300. The CaMV35S::mGFP vector (pCAMBIA1300) was used as a control. The resulting in-frame fusion construct was transiently expressed in onion epidermal cells using a biolistic PDS-1000/He particle delivery system (Bio-Rad, http://www.biorad.com/). GFP signals were observed using a confocal microscope (Zeiss LSM 510, http://www.zeiss.com/).

OsORC3 knockdown by artificial microRNA interference

To suppress OsORC3 expression, artificial microRNA was introduced into pCAMBIA1300 under the control of a CaMV 35S promoter. Primers used to generate miRNA were designed according to the Web MicroRNA Designer (WMD) database (version 2, http://wmd.weigelworld.org/cgi-bin/mirnatools.pl) (I-miR-s, 5'-agtttccgataattcactcgccacaggagattcagtttga-3′; II-miR-a, 5′-tgtggcgagtgaattatcggaaactgctgctgctacagcc-3′; III-miR-s, 5′-cttggcgtgtgtattatcggaaattcctgctgctaggctg-3′; IV-miR-a, 5′-aatttccgataatacacacgccaagagaggcaaaagtgaa-3′). First, three fragments of OsORC3 were amplified by three independent PCR reactions. Second, the three fragments were mixed together and used as templates to generate a stem-loop DNA fragment by fusion PCR, using primers G-4368 and G-4369. The generated 554 bp fragment was then cloned into a pUCm-T vector for sequence confirmation. Finally, the 554 bp fragment was cut using BamHI/PstI and ligated into the pCAMBIA1300 vector under the control of a CaMV 35S promoter. Transgenic plants were generated by A. tumefaciens-mediated transformation. T2 generation homozygous seedlings of two independent single-copy-inserted RNAi transgenic lines (RNAi#1 and RNAi#2) were used in this study.

Microscopic analysis

To observe LRPs, the primary roots were stained with methylene blue (Johnson et al., 1996). The primary roots were fixed in FAA solution (formalin/acetic acid/alcohol 10:5:85 v/v/v) at 4°C for at least 24 h. After fixation, LRPs were rinsed for 10 min in water and stained with 0.01% w/v methylene blue. LRPs were observed and photographed using a Leica MZ95 stereomicroscope coupled to a color CCD camera (Leica, www.leica-camera.com/). To analyze LR development, CYCB1;1::GUS staining was observed using a Zeiss Axiovert 200 microscope with a color CCD camera (Zeiss).

GUS assay, Evans blue staining and Feulgen staining

Histochemical GUS analysis was performed as described by Jefferson et al. (1987). Transgenic plant samples were incubated with X-gluc(5-bromo-4-chloro-3-indolyl-beta-D-glucuronide) solution overnight at 37°C. After staining, the tissues were rinsed and observed using a Leica MZ95 stereomicroscope with a color CCD camera (Leica).

To identify dead cells in rice roots, the roots of 7-day-old seedlings grown in a growth chamber at 30°C/24°C were stained with 2% w/v Evans blue. Evans blue staining was performed for 10 min. Roots were washed three times with distilled water 5 min for each time at room temperature and incubated in water for 2 h, and then analyzed immediately using a dissection microscope (Leica MZ FLIII). Representative results for 20 roots are shown in Figure 6(c–f).

To examine the total DNA content in the root meristem, Feulgen staining was performed, according to a slight modification of the method described by Demchenko et al. (2004). Root tips (8–10 mm) were fixed in Carnoy's fixative for at least 24 h. The root tips were rinsed in 1 m HCl, transferred to 1 m HCl pre-warmed to 60°C for 12 min, and stained with Schiff's reagent until the tip of the root showed distinct coloring (approximately 10 min). Then the root tips were rinsed twice for 2 min in rinse solution (50 mm HCl, 0.5% Na2S2O5). The samples were then placed in 45% v/v HOAc (Acetic acid ) and observed under a stereomicroscope. The primary root tips of 20 seedlings were stained. The root tips of all WT roots exhibited staining, but almost no staining was observed in roots of RNAi lines. Representative results are shown in Figure 6(g,h).


This work was supported by the Ministry of Agriculture of China (2011ZX08009-003-005), the Key Basic Research Special Foundation of China (2011CB100300), the Natural Science Foundation of Zhejiang Province (Y3110093), the Fundamental Research Funds for the Central Universities, and the Program for New Century Excellent Talents in University.