Pescadillo plays an essential role in plant cell growth and survival by modulating ribosome biogenesis



Pescadillo (PES) is involved in diverse cellular processes such as embryonic development, ribosomal biogenesis, cell proliferation, and gene transcription in yeast and metazoans. In this study, we characterized cellular functions of plant PES in Nicotiana benthamiana, Arabidopsis, and tobacco BY-2 cells. A GFP fusion protein of PES is predominantly localized in the nucleolus, where its localization requires the N-terminal domain of PES. Silencing of plant PES led to growth arrest and acute cell death. PES interacts with plant homologs of BOP1 and WDR12 in the nucleolus, which are also nucleolar proteins involved in ribosome biogenesis of yeast and mammals. PES, BOP1, and WDR12 cofractionated with ribosome subunits. Depletion of any of these proteins led to defective biogenesis of the 60S ribosome large subunits and disruption of nucleolar morphology. PES-deficient plant cells also exhibited delayed maturation of 25S ribosomal RNA and suppressed global translation. During mitosis in tobacco BY-2 cells, PES is associated with the mitotic microtubules, including spindles and phragmoplasts, and PES deficiency disrupted spindle organization and chromosome arrangement. Collectively, these results suggest that plant PES has an essential role in cell growth and survival through its regulation of ribosome biogenesis and mitotic progression.


Ribosome biogenesis is a fundamental process that is tightly coordinated with cell growth and proliferation. Disruption of this process is linked to cancer and human disease (Ruggero and Pandolfi, 2003; Lempiäinen and Shore, 2009; Zhang and Lu, 2009). The mechanisms of ribosomal RNA processing and ribosome assembly have been well characterized in yeast based on genetic and proteomic studies (Henras et al., 2008; Kressler et al., 2010; Panse and Johnson, 2010; Karbstein, 2011); however, molecular details of ribosome assembly in higher eukaryotes are less clear. Many of the components involved in the process appear to be evolutionarily conserved (Andersen et al., 2002; Scherl et al., 2002). Plant databases contain homologs of yeast and mammalian ribosomal biogenesis factors, but only a few homologs have been analyzed for their functions in planta (Pendle et al., 2005; Horiguchi et al., 2012). Furthermore, the underlying mechanisms of plant ribosome assembly and the regulation of plant ribosome biogenesis in response to metabolic changes and stress conditions are largely unknown.

Pescadillo (PES) is a locus in the zebrafish mutant that shows severe defects in embryonic development, and its homologs are identified in yeast, mice, and human (Allende et al., 1996; Adams et al., 2002; Lerch-Gaggl et al., 2002). Pescadillo is an evolutionarily conserved nucleolar protein that is indispensible for viability of yeast and higher eukaryotes. The pescadillo-mutant zebrafish dies on the 6th day of embryonic development with smaller eyes and brains; while PES−/− mouse embryos fail to proliferate and die before implantation, suggesting an essential function of Pescadillo in embryonic development (Allende et al., 1996; Lerch-Gaggl et al., 2002). Subsequent studies suggest that Pescadillo has a critical role in diverse cellular processes, such as ribosome biogenesis, rRNA processing, cell proliferation control, and DNA replication (Du and Stillman, 2002; Lerch-Gaggl et al., 2002; Lapik et al., 2004; Grimm et al., 2006). Misregulation of Pescadillo has been linked to chromosomal instability and carcinogenesis (Killian et al., 2004; Li et al., 2009).

Pescadillo (PES) homologs from Zinnia elegans and Arabidopsis thaliana, designated ZePES and AtPES (At5 g14520), have been cloned and characterized (Zografidis et al., 2007). A GFP fusion protein of AtPES is localized to the nucleolus, and both ZePES and AtPES can rescue the mutant phenotype of yeast nop7/yph1, which lacks Pescadillo function, indicating functional conservation between yeast and plant Pescadillo proteins. In Arabidopsis, AtPES is predominantly expressed in meristems and actively dividing cells, yet overexpression of AtPES does not result in any visible phenotype (Zografidis et al., 2007). Since in vivo functions of PES in plants have not been studied to date, we chose to investigate characteristics and physiological functions of PES in Nicotiana benthamiana, Arabidopsis, and tobacco BY-2 cells. Our results suggest that PES has an essential role in plant cell growth and survival, by modulating ribosome biogenesis through a functional link with the nucleolar proteins BOP1 and WDR12.


Virus-induced gene silencing of Pescadillo in Nicotiana benthamiana

Pescadillo (PES) is a single gene in Arabidopsis, and there are two homologous genes in N. benthamiana genome (accession numbers in Table S1) according to the draft genome sequence (; Bombarely et al., 2012). Since N. benthamiana is an allotetraploid species, PES appears to be a single gene in each diploid genome of N. benthamiana. Our analyzes show no knock-out mutant of PES in Arabidopsis T-DNA insertion lines. To investigate molecular functions of PES during plant development, we first examined knock-down phenotypes after virus-induced gene silencing (VIGS) of N. benthamiana PES (NbPES) following the protocol described in Appendix S1. We cloned three different fragments of NbPES cDNA into the Tobacco Rattle Virus (TRV)-based VIGS vector, pTV00, and infiltrated N. benthamiana plants with Agrobacterium that contained each plasmid. TRV:NbPES(F), TRV:NbPES(N), and TRV:NbPES(C) contained a 1.8-kb full-length coding region; a 500-bp N-terminal region; and a 501-bp C-terminal region of the cDNA, respectively (Figure S1a). VIGS with all of the TRV:NbPES constructs resulted in growth arrest and abnormal leaf development (Figure S1b). The affected leaves of the TRV:NbPES plants were small, wrinkled and distorted, and contained localized yellow patches. Real-time quantitative RT-PCR using the primers shown in Table S2 revealed significantly lower levels of endogenous NbPES transcripts in the leaves of TRV:NbPES(F), TRV:NbPES(N) and TRV:NbPES(C) VIGS plants, indicating silencing of NbPES (Figure S1c).

Characterization of dexamethasone (DEX)-inducible AtPES RNAi lines of Arabidopsis

To observe the cellular effects of PES down-regulation in Arabidopsis, we generated transgenic Arabidopsis plants (Col-0 ecotype), which each carried an AtPES RNAi construct that had an inverted repeat of a 350 bp AtPES cDNA fragment under the control of a dexamethasone (DEX)-inducible transcription system (Aoyama and Chua, 1997). Inducible RNAi system can be useful to assess functions of the genes that are essential for basic cell function or development. In DEX-inducible AtPES RNAi lines, RNAi was induced either by germinating seeds on Murashige and Skoog (MS) medium that contained 10 μm DEX (+DEX), or by germinating seeds on MS media and then transferring seedlings to (+)DEX media for further growth. As a control, the RNAi seeds were treated with ethanol (−DEX) in a similar way. When sowed on (+)DEX media, shoot growth of seedlings of two independent AtPES RNAi lines (#28 and #38) was immediately arrested after germination, and perished prematurely without true leaf formation (Figure 1a). When the RNAi seedlings were grown on MS medium and transferred to (+)DEX medium for further growth, they exhibited growth arrest, death of the aerial tissues, and swollen root tips (Figures S2a,b and 1b; top). Confocal laser scanning microscopy of the root tips after propidium iodide staining showed swollen and disorganized cells, suggesting significant defects in cell proliferation and expansion (Figure 1b; bottom). Dark-grown AtPES RNAi seedlings were also defective in hypocotyl elongation (Figure 1c). These results suggest that AtPES is critical for early growth and development in plants. The effect of RNAi on AtPES mRNA levels in the light-grown seedlings was determined by real-time quantitative RT-PCR (Figure S2c). After transfer to (+)DEX media for either 3 or 6 days, seedlings of the RNAi lines exhibited significantly reduced AtPES transcript levels as compared to (−)DEX samples, whereas the transcript levels in wild-type Col-0 plants remained unchanged by DEX treatment (Figure S2c).

Figure 1.

Growth arrest and cell death phenotypes induced by PES-silencing.(a) Growth arrest phenotype of Arabidopsis dexamethasone (DEX)-inducible AtPES RNAi lines. Seedlings were grown 7 days on MS media that contained either ethanol (−DEX) or 10 μm DEX (+DEX).(b) Root morphology. The RNAi seedlings (#28) were grown on MS medium and transferred to (−) or (+)DEX medium for 5 more days of growth. On (+)DEX medium, the seedlings exhibited abnormal roots with swollen tips (top). The roots were stained with propidium iodide and observed by confocal laser scanning microscopy (bottom). Scale bars = 100 μm.(c) Growth defects of AtPES RNAi seedlings grown on (+)DEX media for 5 days in the dark condition.(d) ROS (Reactive oxygen species) production. Seedlings were grown on soil for 11 days, then sprayed with either ethanol (−DEX) or 30 μm DEX for 7 days. The leaves were stained with diaminobenzidine (DAB) to visualize H2O2 production.(e) Confocal images of the TUNEL assay. AtPES RNAi plants (#28) were grown on MS media for 7 days, and then transferred to either (–)DEX or (+)DEX media for either 4 or 7 more days of growth. Leaves were counterstained with DAPI. For a positive control, fixed leaf material of (–)DEX samples (ethanol-treated for 7 days) was subjected to the DNase I treatment. Scale bars = 50 μm.(f, g) Evans blue staining of NbPES RNAi BY-2 cells after 1–6 days of either ethanol or DEX treatment (f). Numbers of live and dead cells (stained blue) were counted (g). Scale bars = 25 μm.

According to Kang et al. (1999), transgenic Arabidopsis plants carrying the empty pTA7002 vector (used for DEX-inducible RNAi) showed a range of growth defects when grown on (+)DEX medium, and several lines among them exhibited severe retardation in plant growth. A correlation was found between the transcript levels of GVG (encoding a glucocorticoid-regulated transcription factor) and the phenotypic severity: the lines with severe phenotypes commonly accumulated high levels of GVG transcripts (Kang et al., 1999). Furthermore, DEX concentration in the medium affected the phenotype in the severe lines: increasing DEX concentration progressively aggravated the defects. The phenotype of the AtPES RNAi lines (#28 and #38) during seed germination was similar to that of severe vector lines in Kang et al. (1999), prompting us to examine whether GVG overexpression contributed to the phenotype of the AtPES RNAi plants. We examined the GVG transcript levels using semiquantitative RT-PCR in multiple AtPES RNAi lines that showed mild to severe phenotypes upon DEX treatment (Figure S3a). We found that the GVG transcript level of RNAi-10 line was similar to that of RNAi-28 and RNAi-38 lines, although the RNAi-10 seedlings exhibited only a mild phenotype, accompanied by minimal gene silencing (Figures S3 and S2d). Thus, the phenotypic severity did not seem to correlate with GVG expression levels among these lines. In addition, treatment of the RNAi-28 lines with as low as 0.1 μm DEX resulted in severe growth retardation, regardless of light conditions during seedling growth (Figures S3b,c and 1a,c). However, according to Kang et al. (1999), 0.1 μm DEX treatment caused no phenotype or only a mild phenotype in the severe vector control lines. Taken together, these results support that AtPES silencing is the main cause of the phenotype in the AtPES RNAi-28 and RNAi-38 lines. However, different strategy may be needed for gene silencing to completely exclude a possible effect of GVG overexpression on the phenotype.

Generation of DEX-inducible NbPES RNAi lines of tobacco BY-2 cells

We also generated transgenic tobacco BY-2 cell lines, which carried a DEX-inducible RNAi construct with an inverted repeat of a 300-bp NbPES cDNA fragment. Real-time quantitative RT-PCR analyzes demonstrated that DEX treatment for 2 days significantly reduced tobacco PES (NtPES) transcript levels in the transgenic BY-2 cells (Figure S4a).

Characterization of the cell death phenotypes in PES-deficient plant cells

AtPES RNAi seedlings were grown on soil and sprayed with either ethanol (−DEX) or 30 μm DEX. Upon DEX spraying, lesions started to form spontaneously in leaves of RNAi plants, leading to premature plant death (Figure 1d). Their cell death was accompanied by excessive accumulation of reactive oxygen species, as shown by brown pigment on their leaves after diaminobenzidine (DAB) staining (Figure 1d). DAB staining visualizes hydrogen peroxide production (Thordal-Christensen et al., 1997).

To examine DNA fragmentation as a programmed cell death (PCD) marker, we used the terminal dUTP nick end-labeling (TUNEL) assay, an assay that detects single- and double-strand DNA breaks by addition of fluorescent nucleotides to free 3′-termini of the broken DNA (Gavrieli et al., 1992). Confocal laser scanning microscopy revealed a strong fluorescent signal in the epidermal nuclei of the AtPES RNAi plants after 7 days of DEX treatment (DOD), suggesting DNA breaks in the nuclei (Figure 1e). In contrast, TUNEL staining was detected neither in control samples ethanol-treated for 7 days [(−)DEX-7 days] nor in (+)DEX samples at 4 DOD. As a positive control, (−)DEX plants were treated with DNase I, which resulted in DNA breaks in the nuclei (Figure 1e). We conclude that RNAi-induced down-regulation of AtPES induces PCD in Arabidopsis.

DEX-inducible NbPES BY-2 cells also started to die soon after DEX treatment, as shown by Evans blue staining (Figure 1f,g). Despite the strong phenotypes of growth arrest and abnormal leaf development, however, none of the NbPES VIGS N. benthamiana lines [TRV:NbPES(F), TRV:NbPES(N), and TRV:NbPES(C)] did not exhibit cell death symptoms, indicating that young seedlings and cell lines may be more sensitive to PES deficiency.

Subcellular localization of AtPES and its mutants

AtPES contains the Pescadillo N-terminus (PES-N) domain at the N-terminus, followed by the BRCA1 C-Terminus (BRCT) domain and a potential sumoylation site (LKKE in AtPES) (Zografidis et al., 2007; Figure 2a). To determine the subcellular localization of AtPES and its mutants, GFP fusion constructs of full-length AtPES and its mutants [carrying deletion of the PES-N domain (∆PES-N), deletion of the BRCT domain (∆BRCT), or a mutation of the potential sumoylation site, LKKE to LAAE (Sumo-m)] were expressed in N. benthamiana leaves via agroinfiltration. Confocal laser scanning microscopy of the leaf epidermal cells revealed that AtPES:GFP was predominantly localized in the nucleolus, although a faint signal was also detected in the nucleoplasm at longer exposure (Figure 2b). ∆PES-N:GFP, however, exhibited no nucleolar signal, but had diffuse nucleoplasmic distribution, suggesting the importance of the PES-N domain for nucleolar localization of PES in plant cells. Interestingly, ∆BRCT:GFP was strongly detected in the nucleolus and at discrete loci in the nucleoplasm. This result contradicts reports for yeast and mammalian PES, in which deletion of the BRCT domain completely abolishes nucleolar accumulation of PES (Hölzel et al., 2007). Disruption of the potential sumoylation site of PES (Sumo-m:GFP) did not affect its predominant nucleolar localization (Figure 2b). These analyzes collectively demonstrate that the PES-N domain strongly contributes, and the BRCT domain partially contributes, to nucleolar accumulation of PES.

Figure 2.

Subcellular localization of AtPES and its mutants.

(a) Structural organization of AtPES. aa, amino acids.

(b) Subcellular localization of AtPES and its mutants. AtPES:GFP and its mutants were expressed in N. benthamiana leaves via agroinfiltration. The infiltrated leaves were briefly stained with DAPI to mark a nucleus and examined by confocal laser scanning microscopy.

Bimolecular fluorescence complementation (BiFC) analyzes of interactions among AtPES, BOP1, and WDR12

Pescadillo associates with both BOP1 and WDR12 to form the PeBoW complex in mammalian cells (Hölzel et al., 2005). Yeast counterparts of three proteins, Nop7/Yph1 (PES), Erb1 (BOP1), and Ytm1 (WDR12), interact with each other for assembly of the Nop7 subcomplex in yeast pre-ribosomes (Miles et al., 2005; Tang et al., 2008). These three proteins are essential and conserved in eukaryotes, and mainly localized in the nucleolus (Strezoska et al., 2000, 2002; Adams et al., 2002; Lerch-Gaggl et al., 2002; Lapik et al., 2004; Grimm et al., 2006). Plant databases contain Arabidopsis and N. benthamiana homologs of BOP1 and WDR12. BOP1 and WDR12 are single genes in Arabidopsis and in each diploid genome of N. benthamiana (accession numbers in Table S1). Arabidopsis BOP1 and WDR12 contain multiple WD40 repeats, which serve as a scaffold for protein interactions (Figure S5a). GFP fusion proteins of Arabidopsis BOP1 and WDR12 were mainly localized to the nucleolus (Figure 3a). We used BiFC to test for an interaction of AtPES with Arabidopsis BOP1 and WDR12 in plant cells (Figures 3b and S5b,c). Using agroinfiltration, AtPES, BOP1, and WDR12 cDNAs were expressed in combination in N. benthamiana leaves as YFPN- and YFPC fusion proteins for confocal laser scanning microscopy. All of the different combinations of protein expression resulted in strong YFP fluorescence in the nucleolus, indicating that AtPES, BOP1, and WDR12 interact with each other in the nucleolus (Figures 3b and S5b,c). However, coexpression of AtPES:YFPN and YFPC, or YFPN and AtPES:YFPC as a control resulted in no YFP fluorescence, indicating a lack of interaction (Figure S5b).

Figure 3.

Characterization of in vivo interactions among AtPES, BOP1, and WDR12.

(a) Nucleolar localization of BOP1 and WDR12. BOP1:GFP and WDR12:GFP were expressed in N. benthamiana leaves via agroinfiltration. The infiltrated leaves were briefly stained with DAPI and examined by confocal laser scanning microscopy.

(b) Visualization of AtPES interactions with BOP1 and WDR12 using BiFC. For BiFC, AtPES, BOP1, and WDR12 were co-expressed as either YFPN or YFPC fusion proteins in N. benthamiana leaves using agroinfiltration. Protoplasts prepared from the infiltrated leaves were observed for YFP fluorescence.

(c) Co-immunoprecipitation. Protein extracts were prepared from N. benthamiana leaves that simultaneously expressed AtPES:Flag and WDR12:HA fusion proteins (left), AtPES:Flag and BOP1:Myc (middle), or BOP1:Flag and WDR12:HA (right). Extracts were subjected to immunoprecipitation (IP) with anti-HA (left and right) and anti-Myc antibodies (middle), and we detected co-immunoprecipitated AtPES:Flag (left and middle) and BOP1:Flag (right) by immunoblotting (IB) with their corresponding antibodies.

(d) VIGS phenotypes. VIGS of NbPES, NbBOP1, and NbWDR12 all resulted in growth arrest and abnormal leaf development in N. benthamiana.

(e) Real-time quantitative RT-PCR analysis of transcript levels of NbBOP1 and NbWDR12 in VIGS plants. Quantification of the relative transcript levels, compared to the TRV sample, is shown. The β-tubulin mRNA level was used as the control. Each value represents the mean ± standard deviation (SD) of three replicates per experiment. Asterisks denote statistical significance as follows: **P ≤ 0.01.

(f) Co-fractionation of AtPES, BOP1, and WDR12 with ribosome subunits. AtPES:Flag, BOP1:Flag, and WDR12:HA were separately expressed in N. benthamiana leaves. After sedimentation of the ribosomes through a sucrose density gradient, the fractions were analyzed by immunoblotting with anti-Flag, anti-HA, and anti-60S ribosomal protein L10a antibodies. Lanes 1–12 indicate the gradient fractions from top (10%) to bottom (50%).

Characterization of the in vivo interaction of PES with BOP1 and WDR12

To further demonstrate the in vivo interaction of AtPES with BOP1 and WDR12, we conducted co-immunoprecipitation assays (Figure 3c). The following pairs were co-expressed in N. benthamiana leaves by agroinfiltration: Flag-fused AtPES (AtPES:Flag) and hemagglutinin (HA)-fused WDR12 (WDR12:HA) (left); AtPES:Flag and Myc-fused BOP1 (BOP1:Myc) (middle); and BOP1:Flag and WDR12:HA (right). WDR12:HA (left and right) and BOP1:Myc (middle) proteins were immunoprecipitated from the cell extracts of infiltrated leaves with either anti-HA or anti-Myc antibodies. We performed western blotting, first, with either anti-HA or anti-Myc, to detect immunoprecipitated WDR12:HA (left and right) and BOP1:Myc (middle), and then with anti-Flag antibodies to detect the presence of AtPES:Flag (left and middle) and BOP1:Flag (right) as co-immunoprecipitates. AtPES:Flag was co-immunoprecipitated with WDR12:HA and BOP1:Myc, and BOP1:Flag was co-immunoprecipitated with WDR12:HA (Figure 3c). However, co-immunoprecipitation did not take place when WDR12:HA (left and right) and BOP1:Myc (middle) were expressed alone (Figure 3c). These results suggest in vivo interactions occur among AtPES, BOP1, and WDR12.

We examined VIGS phenotypes of N. benthamiana BOP1 (NbBOP1) and WDR12 (NbWDR12) in comparison with NbPES VIGS (Figures 3d and S1). TRV:NbBOP1 and TRV:NbWDR12 VIGS constructs contained a 485-bp NbBOP1 cDNA and a 470-bp NbWDR12 cDNA, respectively. VIGS of NbBOP1 and NbWDR12 both resulted in growth retardation and abnormal leaf development, similar to NbPES VIGS. Real-time quantitative RT-PCR revealed significantly lower levels of endogenous NbBOP1 and NbWDR12 transcripts in the leaves of TRV:NbBOP1 and TRV:NbWDR12 VIGS plants, respectively, indicating silencing of these genes (Figure 3e).

Ribosome association of AtPES, BOP1, and WDR12

To investigate the association of PES, BOP1, and WDR12 with ribosomes in a plant cell, leaf cells that expressed AtPES:Flag, BOP1:Flag, or WDR12:HA were fractionated on a 10–50% sucrose density gradient. After ultracentrifugation, fractions were collected, and immunoblot analysis was performed with anti-Flag and anti-HA antibodies (Figure 3f). As a control for fractionation, another immunoblot analysis was performed with antibodies against the 60S ribosomal protein L10a, which is associated with 60S ribosomal large subunits, 80S monosomes, and polysomes. AtPES, BOP1, and WDR12 were most abundant in the fractions that contained 60S large subunits and 80S monosomes, suggesting ribosomal association of these proteins. Interestingly, WDR12 was also detected in the lighter fractions of the gradient, suggesting that WDR12 may be loosely associated with the PeBOW complex or may be involved in other cellular processes (Figure 3f). Recently, human WDR12 has been identified as a cholesterol-regulatory gene, although molecular mechanism of the regulation is unknown (Blattmann et al., 2013).

Reduced global translation activities and reduced synthesis of mature 25S rRNA

Since PES deficiency resulted in immediate growth arrest, and PES proteins cofractionated with ribosomes, we examined cellular protein translation activity, using 35S-labeled methionine, on WT (wild-type) and AtPES RNAi seedlings (#28 and #38; Figure 4a). The seedlings were grown for 7 days on MS media, and then transferred to (−)DEX and (+)DEX media for 4 more days of growth. The profile of radioactive proteins after brief incorporation of 35S-methionine indicated that nascent protein synthesis was significantly decreased in AtPES RNAi seedlings upon DEX treatment (Figure 4a). We also performed a time-course analysis of protein translation in the AtPES RNAi seedlings (#10 and #28) using the 35S-methionine labeling (Figure S6). Protein translation was progressively reduced in RNAi-28 seedlings following 2, 3, and 4 days of DEX treatment, but the RNAi-10 seedlings that lacked AtPES gene silencing (Figures S2d and S3) did not show reduction of nascent protein synthesis at any time points of DEX treatment (Figure S6). These results support that PES deficiency leads to a strong reduction in protein synthesis.

Figure 4.

Reduced global translation and mature 25S rRNA accumulation.

(a) 35S-methionine labeling. Seedlings were grown on MS media for 7 days, and transferred to either (−)DEX or (+)DEX media for 4 more days of incubation. After 35S-methionine labeling of the seedlings, total protein extracts were separated by SDS-PAGE, and the gels were dried and analyzed with a phosphorimager. A duplicate gel was Coomassie-stained to show its Rubisco large subunits (RbcL) as the loading control.

(b) Metabolic rRNA labeling. Seedlings were grown on MS media for 7 days, and then transferred to either (−)DEX or (+)DEX media for either 2 (DEX-2d) or 4 (DEX-4d) more days of incubation. After 32P-UTP labeling of the seedlings, total RNA were separated by agarose gel electrophoresis, transferred to nylon membranes, and analyzed with a phosphorimager. A duplicate gel was stained with ethidium bromide (EtBr) to show the total accumulated mature rRNAs. 35S pre-rRNA, and mature 25S and 18S rRNAs are marked.

(c) Relative 25S/18S rRNA ratio. Band intensities of 25S rRNA and 18S rRNA in the DEX-2d samples shown in (b) were compared.

Pescadillo is required for rRNA processing, and its mutation leads to defective synthesis of 25S/28S rRNA in yeast and mammals (Adams et al., 2002; Lerch-Gaggl et al., 2002; Hölzel et al., 2007). Ribosomal RNA genes in higher plants are organized in large clusters with each gene encoding the 18S, 5.8S, and 25S rRNAs (Rogers and Bendich, 1987). The 18S, 5.8S, and 25S rRNAs are transcribed as a large precursor (35S pre-rRNA) that undergoes complex post-transcriptional processing to produce mature rRNAs (Brown and Shaw, 1998). The rRNA processing pathway in plants is poorly understood, but it likely contains steps that are conserved among eukaryotes (Brown and Shaw, 1998; Zakrzewska-Placzek et al., 2010). However, several events including the order of processing steps and the cleavage sites may be different in higher plants as recently described (Sáez-Vasquez et al., 2004; Comella et al., 2008). We examined the impact of AtPES depletion on nascent synthesis of rRNA molecules by using in vivo [α-32P]-UTP labeling (Figure 4b). With this method, [α-32P]-UTP was incorporated into nascent rRNA transcripts and resulted in three main bands corresponding to 35S pre-rRNA, mature 25S rRNA, and mature 18S rRNA. After transfer to (−)DEX and (+)DEX media for 2 days, the levels of the 35S rRNA precursor were similar in all samples, indicating that transcription of the rRNA genes was not affected by AtPES deficiency at this stage. Nascent synthesis of mature 25S rRNA, however, was significantly reduced in the DEX-treated AtPES RNAi seedlings (Figure 4b). Synthesis of mature 18S rRNA was much less affected. Figure 4(c) presents the relative band intensity ratio of 25S rRNA to 18S rRNA shown in Figure 4(b). The 25S/18S rRNA ratio clearly shows reduced accumulation of mature 25S rRNA in DEX-treated RNAi samples, suggesting inefficient rRNA processing in the AtPES-deficient nucleolus. Consistent with this conclusion, rRNA processing intermediates were moderately accumulated in the RNAi seedlings that were DEX-treated for 2 days, based on real-time quantitative RT-PCR (Figure S7). This is consistent with previous findings that Pes1-, BOP1-, and WDR12-deficient yeast and mammalian cells accumulate only very small amounts of rRNA precursors (Adams et al., 2002; Strezoska et al., 2002; Hölzel et al., 2005). Interestingly, RNAi seedlings incubated for 4 days on (+)DEX medium showed defects in nascent synthesis of both the 35S pre-rRNA and the mature 25S and 18S rRNAs, although the seedlings did not exhibit any visible cell death symptoms at this stage (Figure 4b). These results suggest that further perturbation of nucleolar function started to affect both rDNA transcription and overall rRNA processing. Finally, further treatment with DEX resulted in a decrease in total cellular 25S and 18S rRNAs, suggesting rRNA degradation (Figure S8).

Defective ribosome biogenesis in NbPES-, NbBOP1-, and NbWDR12-silenced N. benthamiana plants

To investigate a potential role of PES, BOP1, and WDR12 in ribosome biogenesis, we performed ribosome profiling in N. benthamiana VIGS plants. Cell extracts were prepared from TRV, TRV:NbPES, TRV:NbBOP1, and TRV:NbWDR12 leaves and fractionated with sucrose density ultracentrifugation, and the resulting ribosome profile was measured at 254 nm (Figure 5). Peak height, detected by A254 nm, indicated the level of each ribosomal species in each sample. Depletion of NbPES, NbBOP1, and NbWDR12 each resulted in severe consequences on ribosomal species. Compared to the TRV control, accumulation of 60S large subunits was significantly reduced in cells, while accumulation of 40S small subunits increased (Figure 5). The amount of 80S monosomes was also reduced, probably due to the blocked assembly of functional 60S subunits. In addition, accumulation of polysomes was also disrupted in TRV:NbPES, TRV:NbBOP1, and TRV:NbWDR12 cells, indicating repression of global protein translation activity (Figure 5). In summary, depletion of PES, BOP1, and WDR12 caused defective biogenesis of the 60S ribosome large subunit and a strong reduction in protein synthesis of plant cells.

Figure 5.

Absorbance profiles of ribosomes at 254 nm.

Ribosomes were purified from leaves of TRV, TRV:NbPES, TRV:NbBOP1, and TRV:NbWDR12 lines using ultracentrifugation on a sucrose density gradient. The ribosome profiles of TRV:NbPES, TRV:NbBOP1, and TRV:NbWDR12 lines were compared with that of control TRV. Experiments were repeated at least three times, and a representative image is shown.

Alteration of the nucleolar structure in NbPES-, NbBOP1-, and NbWDR12-silenced N. benthamiana plants

It has been known that the formation of the nucleolus requires transcription (Mayer and Grummt, 2005; Yuan et al., 2005). However, the mechanisms of forming a defined structure of the nucleolus from the various nucleolar components are not well understood. Previously, defective PES function resulted in disrupted nucleologenesis during mouse embryo development (Lerch-Gaggl et al., 2002). We compared nucleolar morphology of leaf epidermal cells from TRV:NbPES, TRV:NbBOP1, and TRV:NbWDR12 VIGS plants to the TRV control cells (Figure 6). To mark the nucleolus, NbBTF3:GFP was expressed in leaves of VIGS plants by agroinfiltration. NbBTF3 encodes a N. benthamiana homolog of the β-subunit of the nascent polypeptide-associated complex (NAC) (Yang et al., 2007). NAC is a ribosome-associated factor in close proximity to nascent polypeptide chains for folding of the newly synthesized proteins (Rospert et al., 2002; Koplin et al., 2010). Koplin et al. (2010) recently showed that NAC also has a critical role in ribosome biogenesis. In a plant cell, NbBTF3:GFP is predominantly localized in the nucleolus (Yang et al., 2007). After transient expression of NbBTF3:GFP, we briefly stained the leaves with DAPI to mark the nucleus. Confocal microscopy revealed that the leaf epidermal cells from TRV:NbPES, TRV:NbBOP1, and TRV:NbWDR12 lines frequently exhibited morphological alteration of the nucleoli, including fragmented nucleoli and mini-nucleolus-like forms; while TRV control cells predominantly showed a single nucleolus (Figure 6a,b). Meanwhile, DAPI-stained nuclei appeared to be normal in all of the samples. The abnormal nucleolar morphology was also observed by transmission electron microscopy, using leaf sections of VIGS plants (Figure 6c). Although leaf mesophyll cells of the TRV control almost exclusively exhibited a single, well defined, compact nucleolus, TRV:NbPES lines frequently showed small nucleolar fragments in addition to a larger centrally localized nucleolus. These results indicate that the depletion of NbPES and its subsequent defects on ribosome assembly leads to disruption of nucleolar structure.

Figure 6.

Altered nucleolar morphology.

(a) Confocal images of the nucleoli from leaf epidermal cells of TRV control, TRV:NbPES, TRV:NbBOP1, and TRV:NbWDR12 VIGS lines that expressed NbBTF3:GFP. NbBTF3:GFP was used as a nucleolus marker. Nuclei were visualized with DAPI staining.

(b) Frequency of abnormal nucleolar morphology. Numbers of leaf epidermal cells of the VIGS lines that showed altered NbBTF3:GFP fluorescence patterns were counted (n = 100 for each sample).

(c) Transmission electron micrographs of a mesophyll cell (A–C) and its nucleus (D–F) from TRV control (A, D) and TRV:NbPES(N) lines (B, C, E, F). The arrows indicate mini-nucleolus-like structures (E, F). n, nucleus; nl, nucleolus. Scale bars = 5 μm (A–C); 1 μm (D–F).

Localization of PES during mitosis in tobacco BY-2 cells

Inactivation of mammalian Pes1 results in chromosome segregation defects during mitosis in human cell lines (Killian et al., 2004). Pes1 localizes to the periphery of chromosomes early in mitosis, and during telophase, it becomes associated with pre-nucleolar bodies that are present in nascent nuclei (Lerch-Gaggl et al., 2002). RNAi of Pes1 and BOP1 causes aberrant mitosis, inducing multipolar spindles, hyperploids cells, lagging chromosomes, and disorganized metaphase chromosomes in mammalian cells (Killian et al., 2004). No information on PES function during plant cell mitosis exists; therefore, we assessed the localization of PES during mitosis in tobacco BY-2 cells with immunolabelling (Figure 7a). Transgenic BY-2 cells expressing NbPES:Flag were fixed and double-labeled with both anti-Flag and anti-α-tubulin antibodies. Confocal laser scanning microscopy revealed that red fluorescence of NbPES:Flag was predominantly detected in the nucleolus, but also detected faintly in the nucleoplasm during interphase in BY-2 cells, which is consistent with its distribution in N. benthamiana leaves. During prophase, NbPES was found in the disintegrating nucleolus, faintly found in the nucleoplasm, and weakly associated with pre-prophase bands. From metaphase to telophase, NbPES was distributed along mitotic microtubules including spindles and phragmoplasts (Figure 7a). During cytokinesis, NbPES was mainly localized to newly forming nucleoli and partly localized to phragmoplasts. However, based on the microtubule-binding assay, recombinant NbPES proteins did not bind to taxol-stabilized microtubules in vitro (results not shown), suggesting that association of NbPES with spindles and phragmoplasts may occur through the interaction with a microtubule-binding protein. Collectively, NbPES localization along the spindles and phragmoplasts was significantly different from preferential localization of mammalian Pes1 at the chromosome periphery (Lerch-Gaggl et al., 2002), suggesting the PES function during mitosis may differ between plant and animal cells.

Figure 7.

Localization of PES during mitosis and the effects of its deficiency in tobacco BY-2 cells.

(a) Subcellular localization of NbPES during mitosis. Transgenic BY-2 cells that expressed NbPES:Flag were immunolabelled at various phases with anti-α-tubulin (green) and anti-Flag (red) antibodies, and stained with DAPI (blue) for observation with confocal laser scanning microscopy. The arrows indicate nucleoli. Scale bars = 10 μm.

(b) Immunofluorescent staining for mitotic microtubule arrangement in NbPES RNAi BY-2 cells. The BY-2 cells were stained with anti-α-tubulin antibodies (green) and DAPI (blue) following 24 h treatment with ethanol (–DEX) or DEX for observation with confocal microscopy. (A) Bipolar spindles in control metaphase cells (–DEX); after induction (+DEX), PES-deficient cells showed disorganized or multipolar spindles (B), and anaphase chromosome bridges (C). Scale bars = 10 μm.

(c) The number of mitotic RNAi BY-2 cells after ethanol or DEX treatment. The BY-2 cells were treated with 10 μm DEX for 1 day or with 5 μm DEX for 4 days. Thousand cells were counted in each sample.

(d) Frequency of the mitotic defects in the RNAi BY-2 cells from metaphase to telophase stages (n > 200).

Disrupted spindle organization and anaphase bridge formation during mitosis in PES-deficient BY-2 cells

We monitored the progression of mitosis in the DEX-inducible NbPES RNAi BY-2 cells by confocal laser scanning microscopy (Figure 7b–d). The RNAi cells were treated with ethanol (–DEX) or 10 μm DEX for 24 h and then fixed and labeled with DAPI staining and anti-α-tubulin antibodies. The RNAi cells were also treated with 5 μm DEX for 4 days for immunolabelling. Figure S4(b) showed silencing of the NtPES gene in the RNAi BY-2 cells after DEX treatment. The number of mitotic cells in (+)DEX cells (n = 1000) increased compared with (−)DEX cells, indicating delayed mitotic progression (Figure 7c). Visible defects were observed in (+)DEX cells from metaphase to telophase (Figure 7b). The defective mitotic cells exhibited disorganized or multipolar spindles in metaphase, while chromosomes failed to align properly and instead formed a central mass or spreaded around the spindle clusters. In addition, some (+)DEX cells formed chromatin bridges in later stages. Quantification by confocal microscopy revealed the mitotic defects, including multipolar spindles and anaphase chromosome bridges, in 7.5–9.8% of the (+)DEX cells from metaphase to telophase stages (n > 200) (Figure 7d). (−)DEX samples exhibited these mitotic defects in 1.5–2.8% of the cells in the same stages (n > 200) (Figure 7(d)). These results suggest that PES plays a role in spindle organization and chromosome segregation during mitosis in plant cells.


In this study, we showed that PES is an essential protein for plant growth and development, and its deficiency causes growth arrest and acute cell death. PES interacts with BOP1 and WDR12 to form the PeBoW complex, which is predominantly localized in the nucleolus and associated with ribosome subunits. PES modulates accumulation of mature 25S rRNA, and together with BOP1 and WDR12 regulates the assembly of 60S ribosomal large subunits. These results demonstrate that PES and its associated factors have a critical role in regulation of ribosome biogenesis. In yeast, the Nop7 (PES) complex contains three assembly factors, Nop7, Erb1 (BOP1), and Ytm1 (WDR12). The yeast complex is required for proper processing of 27S pre-rRNA into mature 25S and 5.8S rRNAs (Adams et al., 2002). The nop7 mutant exhibits decreased accumulation of mature 25S rRNA, and disrupted ribosome profiles, including decreased 60S large subunits and 80S monosomes, increased 40S small subunits, and accumulation of halfmer polysomes (Adams et al., 2002). Deletion of either Erb1 or Ytm1 also causes similar defects in yeast (Pestov et al., 2001; Miles et al., 2005). It has been recently shown that Nop7, Erb1, and Ytm1 in yeast assemble into pre-ribosomes in an interdependent manner (Miles et al., 2005; Tang et al., 2008). Mammalian counterparts of those proteins, Pes1, BOP1, and WDR12, are essential proteins that are conserved among eukaryotes. These proteins associate with each other to form the PeBoW complex, and they are required for pre-rRNA processing needed for the synthesis of mature 28S and 5.8S rRNAs and the assembly of 60S ribosomal large subunits (Strezoska et al., 2000, 2002; Lapik et al., 2004; Grimm et al., 2006). Based on our results, the PES/BOP1/WDR12 paradigm observed in yeast and mammals appears to be conserved in plants.

Plant PES has different characteristics from yeast and mammalian PES. The BRCT domain of mammalian Pes1 is critical for nucleolar localization and rRNA processing (Hölzel et al., 2007). Pes1 mutants, with either a deleted BRCT domain or a point mutation of conserved residues, exhibit diffuse nucleoplasmic localization and fail to form the PeBoW complex (Hölzel et al., 2005, 2007). Plant PES has a protein structure similar to that of yeast and mammalian PES. Deletion of the PES-N domain in this study, however, completely abrogated the nucleolar localization of plant PES, leading to a diffuse nucleoplasmic distribution pattern; while deletion of the BRCT domain only partially affected it, leading to PES localization in both the nucleolus and the discrete loci of the nucleoplasm. In addition, plant PES was distributed along mitotic microtubules, such as spindles and phragmoplasts, during mitosis, whereas mammalian Pes1 is associated with the chromosome periphery. Nonetheless, RNAi inhibition of PES in tobacco BY-2 cells resulted in the mitotic phenotype similar to those of PES-silenced animal cells (Killian et al., 2004), including formation of disorganized or multipolar spindles and anaphase bridge formation. Thus, while both plant and animal PES proteins play a role in chromosome segregation during mitosis, their action mechanisms and protein interactions are likely to be different. These differences may suggest unique functions of plant PES. Accumulating data suggest that PES plays a role in diverse cellular processes. PES has been linked to chromosome instability and cancer formation, cell differentiation and migration, and transcription control by directly binding to DNA (Killian et al., 2004; Maiorana et al., 2004; Sikorski et al., 2006; Li et al., 2009). PES also has been implicated with organ development in zebrafish and Xenopus laevis (Gessert et al., 2007; Simmons and Appel, 2012). Our results in this study suggest that plant PES may have dual functions in interphase and mitosis, possibly through interactions with different partners. Deciphering the molecular networks of PES activity and its interactions is likely to be essential for understanding its functions in plant cells.

The underlying mechanisms of PCD activation in PES-deficient plants remain to be resolved. AtPES RNAi seedlings exhibit defective nascent synthesis of mature 25S rRNA at 2 days of DEX treatment; however, at 4 days of treatment, the entire process of rDNA transcription and rRNA processing became perturbed, followed by rRNA degradation in later stages. We do not know what triggers this rapid breakdown of nucleolar activity. With 25S rRNA synthesis repressed from PES deficiency, diminished accumulation of pre-ribosomal particles within the nucleoli may be detrimental enough to cause nucleolar collapse. Alternatively, plant PES may have an additional, albeit unidentified, function that is absolutely required to maintain functional nucleoli, and the lack of the activity may trigger nucleolar collapse, repressing rDNA transcription and activating PCD. Intriguingly, the nucleolus is a stress sensor, and rDNA transcription is actively down-regulated in response to nucleolar stress (Mayer and Grummt, 2005). In mammals, TIF-IA, which is the essential RNA polymerase I (Pol I)-specific transcription factor, has a critical role in the stress-dependent inhibition of rRNA synthesis (Mayer et al., 2005). Under stress conditions, TIF-IA becomes inactivated by phosphorylation, and it is translocated from the nucleolus to the nucleoplasm, abrogating the transcription initiation complex formation on rDNA promoters. Depletion of TIF-IA results in disruption of nucleoli, cell cycle arrest, stabilization of p53, and apoptosis (Yuan et al., 2005). In addition to TIF-IA, many Pol I factors and other nucleolar proteins, such as nucleolin and B23, move from the nucleolus to the nucleoplasm, following stress (Mayer et al., 2005). However, there has been no report linking PES to the control of rDNA transcription.

Recent studies indicate that ribosome biogenesis genes have a role in plant development (Byrne, 2009; Horiguchi et al., 2012). A nucleolar protein, TORMOZ (TOZ), is needed to regulate cell division planes in early embryogenesis (Griffith et al., 2007). toz mutant plants exhibit disrupted embryonic cell division patterning that leads to severe developmental defects, such as multiple shoot meristems and many leaves. In contrast, SLOW WALKER1 (SWA1), also a nucleolar protein, is required for the progression of mitotic division cycles during both female gametogenesis and root growth (Shi et al., 2005). Both TOZ and SWA1 proteins contain WD40 repeats, and they are predicted to function during 18S rRNA processing in plant cells, based on the functions of their homologs in yeast. These results suggest that nucleolar functions may interact with regulated cell division processes. OLIGOCELLULA2 (OLI2) encodes Nop2 RNA methyltransferase, which is required for pre-rRNA processing and biogenesis of the ribosomal large subunit (Hong et al., 1997; Fujikura et al., 2009). The oli2 mutant exhibits pointed leaves with reduced palisade mesophyll cell numbers. The phenotypic differences of these mutations in several ribosome biogenesis genes suggest that these proteins participate in specialized nucleolar functions that interact with the cellular control of developmental pathways, in addition to pre-rRNA processing and ribosome biogenesis. The results in this study suggest that PES may have a mitotic function independent of its role in ribosome biogenesis in plant cells. Future studies will uncover the molecular mechanism of cell proliferation control and cell death control by PES and its associated ribosome biogenesis factors in a plant cell.

The glucocorticoid-inducible expression system is widely used for gene function analyzes in higher plants, particularly for the genes of which mutations lead to lethality (Sebastian et al., 2009; Geisler et al., 2012; Min et al., 2013). Plants that carry the glucocorticoid-inducible system are typically treated with 10–30 μm DEX (a synthetic glucocorticoid). The glucocorticoid-inducible system employs a glucocorticoid-regulated transcription factor called GVG, a fusion protein of the GAL4 DNA binding domain, the VP16 transactivation domain, and the receptor domain of the glucocorticoid receptor (Aoyama and Chua, 1997). DEX binding to GVG activates the protein, leading to rapid transcriptional induction of a target gene (Aoyama and Chua, 1997). This system is also useful for studying gene functions in a particular tissue or in a particular developmental stage. However, Kang et al. (1999) observed that the glucocorticoid-inducible system itself caused a number of phenotypic changes in some of the transgenic plants carrying the system, and that severity of the phenotypic change depended on GVG transcript levels in the plants and also on DEX concentration. This may be due to that high levels of activated GVG may bind to additional promoter sites, stimulating transcription of off-target genes (Kang et al., 1999). Another possibility is that the potent VP16 transactivation domain sequesters general transcription factors, inhibiting normal transcription. Thus, despite many advantages of the system, the glucocorticoid-inducible system may need to be used with caution, since for some target genes it may be difficult to distinguish between a target gene-specific effect and a combined effect resulting from high GVG expression.

Experimental Procedures

Detailed descriptions of the following procedures are included in the Supporting Methods (Appendix S1): VIGS, Agrobacterium-mediated transient expression, generation of dexamethasone (DEX)-inducible AtPES RNAi lines in Arabidopsis, generation of DEX-inducible NbPES RNAi BY-2 cell lines, BiFC, real-time quantitative RT-PCR, DAB staining, deoxynucleotidyl transferase-mediated dUTP nick end labeling (TUNEL) assay, transmission electron microscopy, co-immunoprecipitation, incorporation of 35S-labeled methionine, sucrose gradient sedimentation, metabolic labeling of rRNA, ribosome profiling, immunofluorescence, and statistical analyzes.


The authors declare no conflict of interest. This research was supported by the Mid-Career Researcher Program (No. 2012047824) from the National Research Foundation and the Cooperative Research Program for Agriculture Science & Technology Development [Project numbers PJ009079 (PMBC) and PJ008214 (SSAC)] from the Rural Development Administration of Republic of Korea.