Retinoblastoma protein regulates cell proliferation, differentiation, and endoreduplication in plants


(fax +82 31 335 8249; e-mail


Retinoblastoma protein (Rb) plays a key role in cell cycle control, cell differentiation, and apoptosis in animals. In this study, we used virus-induced gene silencing (VIGS) to investigate the cellular functions of Rb in higher plants. VIGS of NbRBR1, which encodes the Nicotiana benthamiana Rb homolog, resulted in growth retardation and abnormal organ development. At the cellular level, Rb suppression caused prolonged cell proliferation in tissues that are normally differentiated, which indicates that Rb is a negative regulator of plant cell division. Furthermore, differentiation of the epidermal pavement cells and trichomes was partially retarded, and stomatal clusters formed in the epidermis, likely due to uncontrolled cell division of stomata precursor cells. Rb suppression also caused extra DNA replication in endoreduplicating leaf cells, suggesting a role of Rb in the endocycle. These Rb phenotypes were accompanied by stimulated transcription of E2F and E2F-regulated S-phase genes. Thus, disruption of Rb function in plants leads to ectopic cell division in major organs that correlates with a delay in cell differentiation as well as increased endoreduplication, which indicates that Rb coordinates these processes in plant organ development.


The mammalian retinoblastoma tumor suppressor protein (Rb) regulates cell division, differentiation and apoptosis in specific cell types. During the cell cycle, Rb plays a key role in regulating the G1/S transition by binding to and repressing E2F transcription factors. E2F-binding sites occur in a wide range of genes, including genes involved in cell cycle progression, DNA replication, and DNA repair (Dyson, 1998; Müller et al., 2001; Weinberg, 1995). In addition to its role in cell cycle control, Rb also promotes the differentiation of many cell types by interacting with other transcription regulators such as MyoD, C/EBPs, and HBP1 (Chen et al., 1996; Lipinski and Jacks, 1999), and it protects cells from apoptosis (Chau and Wang, 2003). In mammals, Rb is inactivated by various mechanisms, including phosphorylation by cyclin-dependent protein kinases (CDKs) during cell cycle progression, and by interacting with viral oncoproteins, which disrupts the binding between Rb and its normal partners (Dyson, 1998; Weinberg, 1995).

In plants, Rb homologs (RBRs) identified in maize and other species show marked domain conservation with their animal counterparts and similar protein-binding characteristics (Ach et al., 1997; Grafi et al., 1996; Huntley et al., 1998; Nakagami et al., 2002; Xie et al., 1996). Increasing evidence suggests that the E2F-Rb pathway plays an important role in controlling G1 to S phase transition in plant cells (den Boer and Murray, 2000; Shen, 2002). In this pathway, CycD/CDK inactivates Rb by hyperphosphorylation leading to dissociation of Rb from the E2F-DP complex, which results in E2F activation and the subsequent transcriptional activation of E2F-controlled S-phase genes (Ramirez-Parra et al., 2003). That the E2F-Rb pathway is important in plant development has been shown by overexpression of E2Fa-DPa (Kosugi and Ohashi, 2003; de Veylder et al., 2002) and cyclin D (CycD)3 (Dewitte et al., 2003). Overproduction of E2Fa-DPa or CycD3 induced uncontrolled cell proliferation and delayed differentiation, and as a result severely affected plant development.

However, there are still limited data to reveal Rb function in cell division, differentiation and/or cell death during plant growth and development. It has been shown that particle bombardment of tobacco BY2 cell cultures with the maize ZmRb1 gene decreased cell division, while overexpression of RepA, a viral replication-associated protein that interacts with and sequesters RBR from its physiological partners, stimulated cell division in the tobacco cultures and increased the frequency of transformation and the callus growth rate in maize (Gordon-Kamm et al., 2002). More recently, it has been revealed that mutant alleles of the gene expressing the Arabidopsis RBR1 protein are gametophytic lethal (Ebel et al., 2004). Mature unfertilized mutant megagametophytes fail to arrest mitosis and undergo excessive nuclear proliferation in the embryo sac, and the central cell initiates autonomous endosperm development, thus revealing a novel function of Rb in gametogenesis and endosperm development (Ebel et al., 2004).

We attempted to address the cellular function of NbRBR1 encoding a Rb homolog in Nicotiana benthamiana, by generating a reduced-expression mutant of NbRBR1 using virus-induced gene silencing (VIGS). This approach circumvents the lethal effect of complete loss of Rb function. Inhibition of Rb function induced prolonged cell proliferation coupled with delayed cell differentiation in the leaves and stems, indicating that Rb positively regulates the cell cycle exit. Rb suppression also enhanced endoreduplication in leaf cells. These cellular phenotypes correlated with transcriptional induction of E2F and E2F-regulated genes. As the NbRBR1 VIGS plants showed retarded growth and abnormal leaf development, we propose that Rb function in the control of cell division, differentiation, and endoreduplication is critical for plant organ development.


Cloning of NbRBR1 cDNA

The partial NbRBR1 cDNA used in the initial VIGS screening was 0.85 kb in length. Its full-length form was obtained by RT-PCR using the sequence information of tobacco NtRb1. The full-length NbRBR1 cDNA encodes a polypeptide of 1003 aa that corresponds to a molecular mass of 111 678.61 Da. The predicted NbRBR1 protein structure shares a high degree of conservation with the RBR proteins of other plant species (Figure 1a), including the conserved domains such as an N-terminal leucine-rich domain, pocket A and B domains, and putative CDK-phosphorylation sites (Ach et al., 1997). NbRBR1 was expressed in all tissues examined, namely, the roots, stems, open flowers, flower buds, and young and mature leaves (data not shown).

Figure 1.

Virus-induced gene silencing (VIGS) phenotypes and suppression of endogenous NbRBR1 transcription.
(a) Schematic depiction of the structure of the NbRBR1 protein-coding region, including the two conserved pocket domains (A and B), the N-terminal leucine-rich region (N), and the putative CDK phosphorylation sites (marked with asterisks). The NbRBR1 cDNA regions used in the VIGS constructs are marked by bars. The positions of the Rb1-N and Rb1-C primer sets that were used in RT-PCR analyses are also shown.
(b) VIGS phenotypes of the whole plants and leaves of the TRV and TRV:Rb1 VIGS lines.
(c) Semiquantitative RT-PCR analyses of the endogenous NbRBR1 mRNA levels in the leaves of the three different TRV:Rb1 VIGS lines. Rb1-N and Rb1-C primers were used. As a control for the RNA amounts, the actin mRNA levels were examined.

VIGS of NbRBR1 and suppression of endogenous NbRBR1 transcription

Virus-induced gene silencing is based on the phenomenon that gene expression is suppressed in a sequence-specific manner by infection with viral vectors carrying host genes (Waterhouse et al., 2001). This approach led us to the finding that NbRBR1 gene silencing resulted in severe growth retardation and abnormal leaf development. To confirm NbRBR1 gene silencing, we cloned three different NbRBR1 cDNA fragments into the TRV-based VIGS vector pTV00 and infiltrated N. benthamiana plants with Agrobacterium containing each plasmid (Figure 1a). TRV:Rb1(N) and TRV:Rb1(C) contain the 1.8 kb N- and 1.2 kb C-terminal regions of the NbRBR1 cDNA, respectively, whereas TRV:Rb1(F) contains the full-length cDNA. VIGS with these constructs all resulted in the same phenotype of abnormal plant development (Figure 1b). This abnormal phenotype has been observed reproducibly in all of the >200 N. benthamiana plants that have been subjected to NbRBR1 VIGS to date. Shoot apical growth was arrested, and the newly emerged leaves were small, irregular-shaped and curled, making a cluster near the shoot apex. Furthermore, flower formation was severely retarded in the NbRBR1 VIGS lines, indicating a role of the Rb in flower development.

The effect of VIGS on endogenous NbRBR1 mRNA levels was examined by semiquantitative RT-PCR (Figure 1c) because the endogenous transcript levels in the leaves are low. RT-PCR using the Rb1-N primers showed significantly reduced PCR product levels in the TRV:Rb1(C) lines relative to the TRV lines, indicating that the endogenous level of NbRBR1 transcripts is greatly reduced in those plants. The same primers also detected high levels of the viral genomic transcripts containing the N-terminal region of NbRBR1 in the TRV:Rb1(N) and TRV:Rb1(F) lines. In turn, the Rb1-C primers showed suppression of endogenous NbRBR1 transcription in the TRV:Rb1(N) lines and detected the viral genomic transcripts in the TRV:Rb1(C) and TRV:Rb1(F) lines. The actin transcript levels, which serve as a control, remained constant (Figure 1c). These results demonstrate that the endogenous NbRBR1 expression was significantly reduced in the VIGS lines and the aberrant phenotypes observed in these lines were caused by suppression of NbRBR1.

Prolonged cell proliferation in the leaves and stems of the TRV:Rb1 lines

The leaves of the TRV:Rb1 plants were small, curled, and distorted (Figure 1b). Transverse leaf sections revealed that while the TRV control leaves had the typical leaf structure of dicotyledonous plants (Figure 2a), the TRV:Rb1 leaves exhibited significantly increased cell numbers and reduced cell sizes in every layer, although the typical dorsoventral organization of the palisade and mesophyll cells was mostly maintained (Figure 2b). Transverse stem sections revealed that the TRV:Rb1 stems also had increased numbers of small-sized cells in every tissue layer except the pith (Figure 2d) relative to the TRV control stems (Figure 2c). Interestingly, large electron-dense nuclei were observed in some epidermal, mesophyll, and palisade cells (Figure 2f, cf. control in Figure 2e), and in some cells in the vascular tissues of the stems in the TRV:Rb1 lines (Figure 2h, cf. control in Figure 2g). The epidermal cells of the TRV lines differentiated into puzzle-shaped pavement cells and stomata but the epidermal cells of the TRV:Rb1 lines were small in size and irregular in shape with abnormal stomatal development (Figure 2i). The cell numbers in the leaf epidermis of the NbRBR1 VIGS line increased approximately threefold compared with the control (Figure 2j). The extra cells in the TRV:Rb1 lines probably arose from additional cell divisions due to a delay in the onset of cell differentiation. Consistently, the mesophyll protoplasts in the TRV:Rb1 leaves were significantly smaller than those in the TRV control (Figure 2k). The chloroplasts and mitochondria in the mesophyll protoplasts from TRV and TRV:Rb1 lines were visualized by chlorophyll autofluorescence and TMRM staining, respectively, followed by confocal laser scanning microscopy. TMRM is a lipophilic cation that accumulates in the mitochondria in proportion to the mitochondrial membrane potential (Zhang et al., 2001). Normal chlorophyll autofluorescence and TMRM staining indicates that the organelles in the leaf mesophyll cells of the TRV:Rb1 lines develop normally (Figure 2k).

Figure 2.

Prolonged cell proliferation in the leaves and stems of the TRV:Rb1 lines. The fourth leaf above the infiltrated leaf and the stem to which the fourth leaf was attached were used.
(a, b) Light micrographs of transverse sections through the central part of the leaf from the TRV (a) and TRV:Rb1 (b) lines.
(c, d) Light micrographs of transverse sections of the stem from the TRV (c) and TRV:Rb1 (d) lines.
(e–h) Light micrographs of the leaf and stem sections of the TRV (e, g) and TRV:Rb1 (f, h) lines. Large nuclei observed in the leaf and stem cells in (f) and (h), respectively, are marked with arrows.
(i) Light micrographs of the abaxial leaf epidermis and fluorescent micrographs of the DAPI-stained epidermis of the TRV control and the TRV:Rb1 lines.
(j) Cell numbers in the leaf epidermis. Data points represent mean ± SD of cell numbers/area (mm2) of the abaxial leaf epidermis of three independent TRV and TRV:Rb1 lines.
(k) Confocal laser scanning micrographs of chloroplasts and mitochondria in protoplasts isolated from the leaves of the TRV and TRV:Rb1 lines. The chloroplasts and mitochondria in the protoplasts were visualized by chlorophyll autofluorescence and TMRM staining, respectively. The false color (blue) was used for chlorophyll autofluorescence to distinguish it from the red fluorescence of TMRM. The merged images of TMRM and chlorophyll autofluorescence and bright-field images are also shown. Scale bars in (a) to (k) = 100 μm.

Clustered stomata and abnormal trichomes in the leaf epidermis of the TRV:Rb1 lines

Stomata are plant epidermal structures that regulate gas exchange. Both the number and distribution of stomata are regulated during leaf development (von Groll and Altmann, 2001). Stomata are formed after a series of asymmetric divisions of transiently self-renewing precursors termed meristemoids. In almost all plant species, stomata are separated from each other by at least one epidermal cell. Scanning electron microscopy revealed that in TRV:Rb1 leaves, a large number of stomatal clusters developed on the uneven surface of the abaxial epidermal layer (Figure 3). These clusters varied in size and contained varied numbers of stomata. Some clusters showed piles of undifferentiated cells around several mature stomata in the center, indicating uncontrolled cell division from meristemoids (Figure 3d,h). These overproduced cells sometimes caused stomatal clusters to bulge out of the plane of the epidermis (Figure 3d,h). The mature stomata in the clusters mostly had normally sized and shaped pores, and the guard cell shapes were mostly normal. The stomata within the clusters were arranged at many angles with respect to each other, and some were in direct contact with one another (Figure 3d,f,h, cf. control in Figure 3c,e,g). Interestingly, the stomatal clusters seemed to be distributed in a pattern that resembled the control, with each cluster positioned equivalent to a single stomata in the control. These phenotypes indicate that Rb suppression primarily affects the number and orientation of cell divisions of the stomata precursor cell in an individual stomatal cell lineage, rather than affecting the initiation of the stomatal pathway.

Figure 3.

Scanning electron micrographs of the leaf epidermis of the TRV and TRV:Rb1 lines. The fourth leaf above the infiltrated leaf from the TRV (a, c, e, g) and TRV:Rb1 lines (b, d, f, h) was used. Trichomes in the TRV:NbRb1 lines (b) appeared to be smaller in size and less abundant than those in TRV control (a). Stomatal clusters formed on the epidermis of TRV:Rb1 lines are indicated by arrows in (d). Trichomes developmentally arrested at the premature stages are indicated by asterisks in (f). (g, h) Magnified views of a single stomata in TRV and a stomatal cluster in the TRV:Rb1 line, respectively. Scale bars = 200 μm in (a) and (b), 100 μm in (c) and (d), 50 μm in (e) and (f), and 20 μm in (g) and (h).

Trichomes that developed on the leaf epidermis of the TRV:Rb1 lines were reduced in number and size compared with those of TRV control (Figure 3b, cf. control in Figure 3a), and some were developmentally arrested at the premature stage (Figure 3f). In a number of species including tobacco and Arabidopsis, trichomes develop very early before the cessation of epidermal cell division (Glover et al., 1998; Marysia et al., 1992; Sachs, 1996; Schnittger and Hülskamp, 2002). These observations indicate that suppression of Rb delays the differentiation of the trichome cells.

Enhanced endoreduplication in the leaf cells of the TRV:Rb1 lines

Endoreduplication often occurs in cells that undergo specialized differentiation, such as hair cells and xylem cells, or in cells that have high metabolic activity such as the endosperm (Joubès and Chevalier, 2000; Sugimoto-Shirasu and Roberts, 2003). Endoreduplicating cells replicate chromosomal DNA without mitosis, thus resulting in a higher-ploidy nucleus. In leaves, endoreduplication takes place only after the cessation of normal mitotic cycles, and thus is a marker for the differentiated state of cells (Joubès and Chevalier, 2000; Sugimoto-Shirasu and Roberts, 2003). Microscopic analyses revealed that some cells in the leaves and stems of the NbRBR1 VIGS lines appeared to contain large nuclei (Figure 2f,h). Using flow cytometry, we measured the ploidy levels in the protoplasts isolated from the mature leaves of the TRV and TRV:Rb1 lines (Figure 4). In three independent TRV control lines, almost all nuclei of the protoplasts (mostly consisting of mesophyll cells) had a ploidy level of 2C, consistent with a previous report on N. tabacum (Kosugi and Ohashi, 2003). However, the protoplasts from the three independent TRV:Rb1 lines exhibited increased numbers of nuclei with a ploidy level of 4C; some nuclei even reached 8C (Figure 4). The increased ploidy levels and large nuclei in the NbRBR1 VIGS lines suggest Rb may be a critical regulator of the endocycle.

Figure 4.

DNA ploidy level in the leaf cells of the TRV and TRV:Rb1 lines.
Protoplasts isolated from the fourth leaf above the infiltrated leaf from each VIGS line were stained with DAPI and analyzed by flow cytometry.
(a) Ploidy distribution of leaf protoplasts from three independent TRV and TRV:Rb1 lines.
(b) Quantification of the results shown in (a).

Upregulation of cell cycle-regulated genes

The hyperproliferative phenotype of the NbRBR1 VIGS lines prompted us to examine the expression of cell cycle-regulated genes by semiquantitative RT-PCR analyses. Transcription of E2F and S-phase genes such as ribonucleotide reductase (RNR), proliferating cell nuclear antigen (PCNA), mini chromosome maintenance (MCM), histone H1, and replication origin protein (CDC6) was significantly upregulated in the Rb-suppressed lines (Figure 5). RNR, MCM, CDC6, and PCNA gene transcription is regulated by E2F in plants (Chaboute et al., 2000; Egelkrout et al., 2002; de Jager et al., 2001; Stevens et al., 2002). Interestingly, the mRNA levels of the phytocalpain-encoding gene NbDEK that plays a critical role in cell proliferation and differentiation during plant organ development (Ahn et al., 2004) were also elevated in the TRV:Rb1 lines (Figure 5). Thus, the prolonged cell proliferation in the TRV:Rb1 tissues correlates with E2F activation and increased expression of S-phase genes, which probably have led to the initiation of DNA replication and cell cycle progression. There is also a possibility that loss of proper differentiation signals due to inhibited Rb function caused the increased cell proliferation phenotype and overexpression of cell division genes.

Figure 5.

Expression of cell cycle-related genes.
Transcription of cell cycle-related genes was determined by semiquantitative RT-PCR. TRV control and two independent TRV:Rb1 lines were analyzed. The fourth leaf above the infiltrated leaf from each virus-induced gene silencing line was used. As a control, the transcript level of actin was examined.


In animals, the proper execution of various differentiation pathways requires cell cycle exit (Guo and Walsh, 1997; Zhang et al., 1998). The E2F-Rb pathway plays a critical role in coordinating proliferation and differentiation by regulating the G1/S transition during the cell cycle (Goodrich et al., 1991; Weinberg, 1995). Recent findings suggest that the mechanisms that control the G1/S transition in plants are similar to those in mammals. ZmRb1 interacts with CycD through a conserved A/B pocket domain (Huntley et al., 1998), and NtRb1 is phosphorylated by the CycD/CDK complex at the G1/S-phase transition in tobacco cells (Nakagami et al., 2002). Furthermore, plant E2Fs interact with plant RBR proteins and function as transactivators (de Jager et al., 2001; Mariconti et al., 2002), regulating the expression of genes required for DNA replication and cell cycle progression (Chaboute et al., 2000; Egelkrout et al., 2002; de Jager et al., 2001; Ramirez-Parra et al., 2003; Stevens et al., 2002). Under normal circumstances, the expression of E2F-dependent genes is inhibited by the formation of a complex between E2F and Rb. In this study, reduction in the cellular levels of Rb indeed resulted in activated expression of various S-phase genes, indicating release of E2F factors from Rb-mediated transcriptional repression. Rb suppression also led to increased levels of the E2F mRNA. This is probably also caused by the activation of E2F as activation of E2Fs in mammalian cells induces the expression of the CycD3, E2F1, and E2F2 genes (Müller et al., 2001). Also consistent with our findings is that the Rb null mutation in the mouse induced the ectopic expression of S-phase genes such as histone H1, MCM, PCNA, and RNR (Black et al., 2003). Interestingly, we found that reduced Rb activity also upregulates the expression of the calpain homolog-encoding gene NbDEK (Ahn et al., 2004; Lid et al., 2002). NbDEK controls plant organogenesis by regulating cell proliferation and differentiation, and it appears to act partly by controlling the E2F-Rb pathway (Ahn et al., 2004). That NbDEK is upregulated in our NbRBR1 VIGS plants supports the possibility that NbDEK and the E2F-Rb pathway interact.

Rb is established as the most crucial regulator of E2F activity. Consistently, the phenotype of Rb suppression observed in this study largely overlaps with the phenotypes of E2Fa-DPa-overexpressing Arabidopsis (de Veylder et al., 2002) and tobacco (Kosugi and Ohashi, 2003). VIGS-induced Rb suppression resulted in plants with more cells than the control plants due to a prolonged proliferation phase that correlated with E2F activation and the uncontrolled expression of cell cycle genes. Cell differentiation was concomitantly delayed, indicating that Rb is required for the inactivation of E2F for the cell cycle exit, and possibly for activation of cell differentiation pathways. These combined defects due to Rb depletion caused abnormal plant development, which demonstrates that the correct balance between cell division and differentiation is essential for plant development. Similarly, E2Fa-DPa overexpression in Arabidopsis and tobacco resulted in arrested plant development at an early stage and induced sustained cell proliferation in normally differentiated tissues; this correlated with the transcriptional induction of S-phase genes, including the preRC (pre-replication complexes) genes (Kosugi and Ohashi, 2003; de Veylder et al., 2002). The E2Fa-overexpressing Arabidopsis plants also showed delayed cell differentiation into the epidermal pavement cells and the abnormality in stomata development (de Veylder et al., 2002), which further resembles the phenotype caused by Rb suppression that we observed in this study. These similarities suggest that the major function of Rb in plant cells is to control E2F-DP transcriptional activity. Interestingly, apoptosis-like cell death was induced neither by Rb suppression (in this study) nor by E2Fa-DPa overexpression in plants (Kosugi and Ohashi, 2003; de Veylder et al., 2002). This contrasts with the findings in animal systems that showed loss of Rb or E2F overexpression induced intensive apoptosis (Dyson, 1998; Lipinski and Jacks, 1999).

The enhanced endoreduplication observed in the leaves of both the NbRBR1 VIGS lines and the E2Fa-DPa-overexpressing transgenic plants (Kosugi and Ohashi, 2003; de Veylder et al., 2002) indicates that the E2F-Rb pathway plays a critical role in regulating the endocycle, although the mechanism involved remains unclear. That overexpression of CDC6, one of the E2F-responsive genes, induces endoreduplication in Arabidopsis (Castellano et al., 2001) suggests that the CDK-activated E2F-Rb pathway promotes endoreduplication, at least in part through the transcriptional activation of the preRC proteins. A possible scenario that explains the elevated endoreduplication in the NbRBR1 VIGS leaves is that some cells may lack a mitosis-inducing factor because of progression into a differentiated state, and then the sustained E2F activity because of Rb depletion may activate these cells to re-enter the S-phase, thus resulting in increased ploidy levels. Consistent with our finding is that hyperphosphorylation of Rb by S-phase CDKs correlates with endoreduplication in maize endosperm development (Grafi et al., 1996). Interestingly, CycD3 overexpression in the whole plant and trichomes of Arabidopsis resulted in reduced endoreduplication in leaves and trichomes, respectively (Dewitte et al., 2003; Schnittger and Hülskamp, 2002). That is intriguing, considering that E2F-DP acts downstream of D-type cyclins in the Rb pathway. Furthermore, CycD3 overexpression in Arabidopsis induced ectopic cell division and delayed cell differentiation resulting in severely defective plant development (Dewitte et al., 2003), a phenotype with some parallels to the E2F-DP overexpression phenotype (de Veylder et al., 2002) and the Rb suppression phenotype (in this study). A possible reason for the differences in the endoreduplication levels is that most of the CycD3-overexpressing cells might have failed to reach the differentiation stage at which endoreduplication is initiated. Alternatively, ectopic expression of CycD3 may directly inhibit endoreduplication, as supported by the recent study that continuous CDK activity inhibits preRC assembly onto replication origins (Edgar and Orr-Weaver, 2001).

Suppression of NbRBR1 caused stomatal clusters to form in the leaf epidermis, dramatically altering the stomatal patterning. Stomatal development begins with an unequal cell division of a meristemoid mother cell (MMC) to produce a meristemoid and a subsidiary cell. After a number of unequal cell divisions, the meristemoid becomes a guard mother cell (GMC). The GMC then undergoes a final, equal cell division that gives rise to the two guard cells that form the stoma (Serna and Fenoll, 2000, 2002). The stomata-free region surrounding each stoma forms when the plane of asymmetric divisions is oriented so that the new precursor, the satellite meristemoid, does not contact the preexisting stoma or precursor. The phenotype of the stomatal clusters in the NbRBR1 VIGS lines (Figure 3) suggests that in these plants, many MMCs apparently divided so that the new meristemoids contact the existing stomata or meristemoids. This indicates that Rb suppression disrupts the orientation of the asymmetric divisions involved in stomatal development. Furthermore, the Rb suppression seemed to promote ectopic cell divisions from meristemoids to increase the number of stomata produced by an individual stomatal cell lineage. In addition, differentiation of the resulting daughter cells from the asymmetric division into pavement cells may also have been delayed by the compromised Rb function. However, the overall distribution patterns of the stomatal clusters on the epidermis of the TRV:Rb1 lines indicate that suppression of Rb did not drastically alter the numbers of cells entering a stomatal cell lineage. Recently, it has been reported that B1-type cyclin-dependent protein kinases are essential for the formation of stomatal complexes in Arabidopsis (Boudolf et al., 2004). Reduced B-type CDK activity caused an early block of meristemoid division and inhibited satellite meristemoid formation. These findings also suggest that the control of cell division is critical for normal stomatal development.

The stomatal clusters observed in this study morphologically resembled those in several Arabidopsis mutants, including the stomata developmental mutant too many mouths (tmm) (Nadeau and Sack, 2002; Yang and Sack, 1995) and the photomorphogenic mutant cop10 (Wei et al., 1994). The TMM gene encodes a leucine-rich-repeat (LRR)-containing receptor-like protein that has a structure very similar to that of CLAVATA2 (Jeong et al., 1999). This indicates that TMM is involved in the cell-to-cell signaling pathway that regulates stomatal patterning (Nadeau and Sack, 2002). COP10 plays important roles in plant growth and development, and stomatal development appears to be affected by the abnormality in those developmental processes. It will be interesting to probe whether Rb is involved in the TMM signaling pathway. The TMM signaling pathway appears to control the plane of patterning division as well as the balance between stem cell renewal and cell differentiation (Geisler et al., 2000; Nadeau and Sack, 2002). It may be that Rb acts downstream of TMM in these processes.

NbRBR1 suppression seemed to affect different cell layers with different severity. NbRBR1 suppression strongly affected the cell-fate determination of the leaf epidermis but had little effect on the cell specificity of the palisade and mesophyll layers (Figure 2). Moreover, in the stems, the hyperproliferation phenotype of Rb was most visible in the epidermis and the vascular tissues while the innermost pith layer was much less affected. The cell type-specific effects of Rb mutation were also seen in the Rb-null mice, in which most cell types appeared to develop normally despite the essential nature of the gene (Jacks et al., 1992; Lee et al., 1992). It is possible that some cell types contain specific components that are able to compensate for the loss of Rb function in plants. In mammals, there are two additional Rb family members, p107 and p130, that have overlapping functions with Rb (Classon and Dyson, 2001). Given that mammalian Rb interacts with various proteins involved in cell cycle control and cell differentiation (Lipinski and Jacks, 1999; Morris and Dyson, 2001), it is possible that plant Rb may also interact with additional pathways and regulators of plant development, distinct from the basic components of G1/S control. Rb interaction with those different cell type-specific regulatory components may have alleviated the effects of Rb mutation in those cells.

Experimental procedures

Virus-induced gene silencing

The NbRBR1 cDNA fragments were polymerase chain reaction (PCR)-amplified and cloned using XcmI sites into the pTV00 vector that contains a part of the TRV genome (Ratcliff et al., 2001). VIGS was carried out as described (Ahn et al., 2004; Kim et al., 2003; Ratcliff et al., 2001).

Semiquantitative reverse transcription (RT)-PCR

Semiquantitative RT-PCR was performed with 5 μg total RNA isolated from the fourth leaf above the infiltrated leaf as described (Ahn et al., 2004). The RT-PCR primers used are as follows: Rb1(F), 5′-atggtggagctgaataat-3′ and 5′-ctaagactcaggctgctc-3′; Rb1(N), 5′-atggtggagctgaataat-3′ and 5′-acatagtctgcagaaccc-3′; Rb1(C), 5′-agaaagtgtgttccgaat-3′ and 5′-ctcaggctgctcagtttt-3′; PCNA, 5′-ggaattacggcttg ttcagg-3′ and 5′-ttcaatcttaggtgagg-3′; MCM, 5′-agaccgccttagctcttgcg-3′ and 5′-tgaggaattatctc agggtc-3′; and histone H1, 5′-gctgtggagtccgatgaacc-3′ and 5′-ctatctattacttcctcccc-3′. The sequences of the remaining primers have been described (Ahn et al., 2004).

Confocal laser scanning microscopy

Leaf protoplasts were stained with 200 nm tetramethylrhodamine methyl ester (TMRM, Molecular Probes, Eugene, OR, USA) for 10 min and TMRM fluorescence and chlorophyll autofluorescence of the protoplasts were observed by using a confocal laser scanning microscopy as described (Kim et al., 2003).

Histochemical analyses

Scanning electron microscopy, tissue sectioning, and light microscopy were carried out as described (Ahn et al., 2004) using the fourth leaf above the infiltrated leaf from the TRV and TRV:Rb1 lines. Plant tissues collected from TRV and TRV:Rb1 lines were sequentially fixed with 2.5% (v/v) glutaraldehyde and 1% osmium tetraoxide, dehydrated by applying an ethanol series, then embedded in Spurr's resin (EM sciences, Gibbstown, NJ, USA). Thin sections were then made with an LKB III ultramicrotome and stained sequentially with 5% uranyl acetate and 3% lead citrate, followed by observation under a transmission electron microscope (model JEOL 1200 EXII; JEOL, Tokyo, Japan). Samples were fixed in FAA solution under a vacuum as described (Ahn et al., 2004). After dehydration by applying a graded ethanol series, the specimens were critical point-dried in liquid CO2. The dried materials were mounted and coated with platinum-palladium in a sputter-coater and examined by a scanning electron microscope (model JSM-6700F; JEOL).

Flow cytometry

Protoplasts isolated from the middle region of the fourth leaf above the infiltrated leaf from three independent plants of the TRV control and TRV:Rb1 lines were stained with DAPI for flow cytometry according to the manufacturer's instructions (Partec, Münster, Germany). The DNA content of the protoplasts was analyzed by using the PASIII flow cytometer (Partec). In total, 14 990, 16 185, 16 398, 14 878, 17 173, and 13 968 protoplasts from the TRV-1, TRV-2, TRV-3, TRV:Rb1-1, TRV:Rb1-2, and TRV:Rb1-3 lines, respectively, were analyzed by flow cytometry.


The authors wish to thank Dr David C. Baulcombe (John Innes Center, Norwich, UK) for providing VIGS vectors. This research was supported by grants from the Molecular and Cellular BioDiscovery Research Program (to H.-S. Pai), the Plant Diversity Research Center of the 21st Century Frontier Research Program (to H.-S. Pai and W.T. Kim), KOSEF Basic Research Program (to H.-S. Pai), the Crop Functional Genomics Center (to J.-K.K.), and the Plant Metabolism Research Center of the Science Research Center Program (to W.T. Kim), all of which are funded by the Ministry of Science and Technology of the Korean Government.

EMBL accession number: AY699399