|Home | About | Journals | Submit | Contact Us | Français|
Protein degradation by the ubiquitin-proteasome system is necessary for a normal cell cycle. As compared with knowledge of the mechanism in animals and yeast, that in plants is less known. Here we summarize research into the regulatory mechanism of protein degradation in the cell cycle in plants. Anaphase-promoting complex/cyclosome (APC), in the E3 family of enzymes, plays an important role in maintaining normal mitosis. APC activation and substrate specificity is determined by its activators, which can recognize the destruction box (D-box) in APC target proteins. Oryza sativa root architecture-associated I (OsRAA1) with GTP-binding activity was originally cloned from rice. Overexpression of of OsRAA1 inhibits the growth of primary roots in rice. Knockdown lines showed reduced height of seedlings because of abnormal cell division. OsRAA1 transgenic rice and fission yeast show a higher proportion of metaphase cells than that of controls, which suggests a blocked transition from metaphase to anaphase during mitosis. OsRAA1 co-localizes with spindle tubulin. It contains the D-box motif and interacts with OsRPT4 of the regulatory particle of 26S proteasome. OsRAA1 may be a cell cycle inhibitor that can be degraded by the ubiquitin-proteasome system, and its disruption is necessary for the transition from metaphase to anaphase during root growth in rice.
Protein degradation by the ubiquitin-proteasome system is necessary for the normal cell cycle. The activation of 3 enzymes, E1 (ubiquitin-activating enzyme), E2 (ubiquitin-conjugating enzyme) and E3 (ubiquitin ligase), are required for the addition of ubiquitin molecules to the target protein. E1 catalyzes the formation of the thiol-ester bond between C-terminal glycine in ubiquitin and cysteine in E1, and activated ubiquitin is transferred to a cysteine in E2. With the help of an E3, ubiquitin is linked to the lysine in the target protein. Subsequent ubiquitins can be attached to the previously bound ubiquitin because of the seven lysine residues in the ubiquitin molecule. Finally, the ubiquitinated substrates are degraded by the 26S proteasome.
E3 confers substrate specificity. E3 ubiquitin ligases comprise a large and diverse family of proteins or protein complexes. E3s are of two classes: homology to E6-AP carboxy terminus-containing proteins, and RING-finger domain-containing proteins. The RING-finger E3s have 4 subgroups: single subunit RING E3, VCB-Cul2 complex (VBC), Skp1/Cullin/F-box protein (SCF) and anaphase-promoting complex/cyclosome (APC/C).1 The SCF ligases regulate the transition from G1/S and G2/M, and APC is required for mitosis. Many APC substrates have been identified in animals.2 The polyubiquitinated substrates can be recognized by different ubiquitin receptors and degraded via 26S proteasome.3,4 However, little is known about APC substrates in plants.
Oryza sativa root architecture-associated 1 (OsRAA1), originally cloned from rice,5 encodes a small GTP binding protein of 12 kD.6 It has a phylogenetic relationship with the reported small G proteins (Fig. 1). Most of the G proteins regulate vesicle trafficking.7 Arf participates in polar auxin transport to regulate root development in rice.8 ROP (Rho-related GTPase from plants) can activate the receptor-like cytoplasmic kinase to switch on downstream signaling.9 In Arabidopsis, ROP1 regulates actin dynamics to control pollen tip growth,10,11 and stomatal aperture may be controlled by G-protein ROP2 activity.12 TaRAN1 protein can regulate G2 cell proportion in the cell cycle to affect the development of lateral roots.13 OsRAC1 localizes in the plasma membrane and mediates innate immunity in rice.14,15 Phylogenetic assay revealed RAA1 close to the known small G proteins such as Pho4 in yeast, Rop3/6 in maize and Rac1/2 in rice, which suggests a possible relation with function in this family.
OsRAA1 is a homologous gene of flowering promoting factor 1 (FPF1), first cloned from mustard.16 They share 58% protein identity. OsRAA1 and FPF1 share the similar motif, RGSLDLISL and RSSLDLISL, respectively. OsRAA1 expression can be induced by auxin, and its overexpression inhibits growth of primary roots, increases the number of adventitious roots and increases the level of endogenous indoleacetic acid in rice. The GTP binding protein of OsRAA1 may regulate root development mediated by auxin.5 In RNAi knockdown transgenic rice lines, the reduced height of seedlings caused abnormal cell division, without altering root development.6 Ectopic expression of AtFPF1 in transgenic rice lines produced root architecture phenotypes similar to those of OsRAA1 transgenic rice.5,17
Our recent experiments indicated that overexpression of OsRAA1 or AtFPF1 caused early flowering in Arabidopsis.18 AtFPF1 promotes flowering in parallel to a constitutive pathway mediated by gibberellin.19 OsRAA1 transgenic Arabidopsis also displayed the gibberellin-sensitive phenotype during germination (our unpublished data). In addition, transgenic Arabidopsis plants overexpressing OsRAA1 or AtFPF1 exhibited reduced sensitivity to red light, which resulted in longer hypocotyls.18 Therefore AtFPF1 and OsRAA1 may have the same functions in promoting flowering and regulating hypocotyl length via the gibberellin pathway in Arabidopsis or in controlling root architecture via the auxin pathway in rice.
The growth and development of plants depends on the cell cycle. The major phases of the cell cycle are G1 (postmitotic interphase), S (DNA synthesis phase), G2 (premitotic interphase) and M (mitosis/cytokinesis) and are governed by the control of cyclin-dependent kinases (CDKs) and cyclin protein.20 CDKs and cyclins are important for the G0/G1, G1/S and G2/M transitions. Approximately 80 genes regulate the cell cycle.21 Although extensive studies have revealed the roles of some cell cycle regulators and the underlying mechanisms in plants,22–24 the function of few cell cycle regulators in rice have been analyzed.25 OsFSM regulates rice development by regulating the duration of the S and G2 phases. The fsm rice mutant showed defective seedling growth and was lethal in the seedling phase.26
OsRAA1 is a novel cell cycle regulator during rice root development.27 OsRAA1 transgenic rice showed a higher proportion of metaphase cells and a lower proportion of anaphase cells than that of the wild type, which suggests that the mitosis process is blocked in the transition from metaphase to anaphase in OsRAA1-overexpressed rice. Immunolocalization assay revealed that OsRAA1-GFP can co-localize with tubulins in spindles during mitosis in tobacco BY2 cells. In fission yeast cells transformed with OsRAA1, more than half of the cells were blocked in metaphase. In addition, OsRAA1 can interact with kinesin-related protein. From observations in transgenic yeast, tobacco and rice, OsRAA1 with GTP binding activity has a conserved function in cell cycle regulation.27
APC can be divided into 4 parts. One scaffolding protein complex contains Apc1, Apc4 and Apc5. The second part is a catalytic core of Apc2, Apc11 and Apc10/Doc1, which interacts with E2 ligase. Apc10/Doc1 may be responsible for substrate recognition. The third part is tetratricopeptide repeat (TPR) arm which consists of two subunits each of Apc8/Cdc23, Swm1, Apc6/Cdc16, Cdc26, Apc9 and Apc3/Cdc27. Apc3/Cdc27, Apc6/Cdc16, Apc7 and Apc8/Cdc23 all contain the protein-protein interaction motif TPR, which mediates protein-protein interaction in multiprotein complexes. Apc13/Swm1, Apc9 and Cdc26 play roles in TPR complex integrity. The fourth part is the adaptor Cdh1.
Arabidopsis was predicted to have 2 E1 genes, 37 E2 genes and ~1,300 genes that encode E3 components.28 APC is one of the 4 E3 subgroups and is well conserved. All of the APC subunits in animals have a counterpart in Arabidopsis.29,30 So far, only a few AtAPC subunits have been reported. AtApc2 interacts with AtApc11 and AtApc8. AtApc2-null mutation caused the accumulation of D-box-containing cyclin.31 Most of the APC subunits are encoded by a single-copy gene. However, AtApc3/Cdc27 is encoded by two genes, AtCdc27a and AtCdc27b. Both HBT/Cdc27b and Cdc27a can interact with the Apc2 subunit in vivo. HBT/Cdc27b can interact with at least Cdc20.1, Cdc20.2, Cdc20.5, Ccs52A1, Ccs52B.32 Overexpression of AtCdc27a in tobacco resulted in increased size of organs and biomass and development rate.33 AtCdc27a has 2 forms derived from the alternative splicing of its pre-mRNA.29 HBT/Cdc27b is localized in mitotic cells, spatially concentrated on the spindle at anaphase, which supports HBT/Cdc27b being required for cell division and expansion.32,34 In addition, 2 forms of Apc7 in Arabidopsis result from pre-mRNA splicing, AtApc7 contains 558 amino acids and AtApc7Spl only C-terminal 276 amino acids. The N-terminal region of AtApc7 is required for interaction with other APC subunits.29 Ovules of the nomega mutant showed accumulation of cyclin B proteins. NOMEGA encodes the homolog of the Apc6 subunit in Arabidopsis.35 Observation of mRNA levels in plant tissues suggested that more than one APC form may function during plant development.29
APC activation and substrate specificity are conferred by different activators: Cdh1 (Ccs52 in plants), Cdc20 (cell division cycle 20) and Ama1 in budding yeast.2,36 These proteins contain seven WD40 repeats and two APC binding motifs, a C-terminal isoleucine-arginine dipeptide (IR) motif, an N-terminal C-box, and cyclin-dependent kinase phosphorylation sites. In vitro, the conserved IR motif of Cdh1 and Cdc20 can interact with the TPR of Apc3 and Apc7.37,38 Recent results indicated that the activators bind to Apc3 subunit and also to Apc8/Cdc23.39 Both Cdc20 and Cdh1 can recognize the D-box in APC substrates.40 Cdc20 and Cdh1 function as adaptors to link the APC core and substrate. They can bind to Cdc27, Apc2 and Cdc23 though three sites: the IR motif, C-box and an unknown motif, respectively.39,41
Cdc20 protein accumulates in the S phase and peaks in level in mitosis. In late anaphase and G1, Cdc20 is targeted and degraded by APCCdh1. APCCdc20 activity is inhibited by early mitotic inhibitor 1 (Emi1) by the spindle assembly checkpoint and protein kinase A signaling. The degradation of Emi1 leads to the release of the inhibitory state of APCCdc20.2,36 The Arabidopsis genome contains five Cdc20 genes.42,43 Yeast two-hybrid assay revealed that at least five activators can interact with HBT/AtCdc27b/AtApc3.32
Cdh1 is another activator for APC. APCCdh1 activity is regulated by Cdh1, Cdh1 phosphorylation, Emi1 and other proteins. The Arabidopsis genome contains three Ccs52 genes.42–44 The transcripts of three AtCcs52 genes, AtCcs52A1, AtCcs52A2 and AtCcs52B, are present at different phases: AtCcs52A1 and AtCcs52A2 transcripts are distributed from late M to early G2, whereas AtCcs52B transcripts are distributed from G2/M to M. Therefore, APC activators act consecutively in plants.40,44 The two isoforms of AtCcs52A regulate root development by different mechanisms. Mutant ccs52a1 plants show longer roots than do wild-type plants, with mutant ccs52a2 plants showing a shorter root. Although the two proteins AtCcs52A1 and AtCcs52A2 are functional homologues, their functions differ depending on their promoter activity in root development. AtCcs52A1 is expressed in the elongation zone to stimulate mitotic exit, and Ccs52A2 is expressed in the root meristem to maintain the mitotic silent state of the quiescent center.45 AtCcs52A2 transcription is restrained by E2Fe directly.46 In Arabidopsis, the presence of 2 APC3/Cdc27, 5 cdc20 and 3 Ccs52 genes confers numerous active forms of APCCdc20 and APCCcs52 depending on their subunit composition, but the promoter specificity confers different localizations of APCCcs52A during plant development.
An estimated 10% of proteins are regulated by the ubiquitin-26S proteasome pathway; thus, ~26,000 target proteins can be ubiquitinated by E3. Two types of E3, SKP2 and APC, are involved in the plant cell cycle.28 The common structural motif of APC substrates, mitotic cyclins, is D-box, with the consensus amino acid sequence RxxLxxxxN. The arginine at position 1 and leucine at position 4 are highly conserved.2,47,48 The D-box can be recognized by an APC activator, Cdc20 or Cdh1. With the variability in the D-box sequence, the different mitotic cyclins can selectively interact with APC activators. To transition from metaphase to anaphase, several cell cycle inhibitors, such as cyclin B and securin, need to be degraded. In metaphase, cyclin B and the inhibitory protein of separase, securin, are ubiquitinated by active APCCdc20 and APCCdh1, respectively.2
In plants, the A- and B-type cyclins and securin are APC substrates and can be recognized in a D-box-dependent manner.30 For example, AtCycA2;3 acts as a negative regulator for endocycles. With mutation of the leucine at position 4 in the D-box, CycA2;3 protein accumulates and promotes mitosis and decreases the ploidy levels.49 In Arabidopsis, the presence of 21 mitotic cyclins, 8 activators and 5 Cdc20 and 3 Ccs52 proteins might be required for APC activity throughout the cell cycle.44 Both the D-box motif and the highly divergent N-terminal sequences are required for the interaction between AtCcs52 and cyclins. Substrates are recognized by the activator and can bind to an unknown site on the APC core.39 The targeted proteins are hydrolyzed by the protein-cutting machine, 26S proteasome.
The 26S proteasome exists in the cytoplasm and nucleus, especially in rapidly dividing tissues.28 It is composed of a 20S core and a 19S regulatory particle (RP).50 The 19S RP is composed of 18 subunits that can be divided into two subcomplexes, the lid and the base. The lid has nine no-ATPase subunits (RPN3, RPN5 to RPN9 and RPN11 to RPN13). The base contains three no-ATPase subunits (RPN1, RPN2 and RPN10) and six AAAATPase subunits (regulatory particle triple-A ATPase subunits RPT1, RPT2, RPT3, RPT4, RPT5 and RPT6).51
The 19S particle is responsible for recognizing and unfolding the target proteins and then transferring the substrates to the 20S chamber for hydrolysis. The functions of some RPN subunits have been explored. Recent work showed that the “two hands”, RPN10 and RPN13, function as primary receptors for ubiquitin to grasp the ubiquitinated substrates.52 In higher plants, most of the subunits in 19S RP are encoded by two or more genes.53 AtRPN1 is encoded by AtRPN1a and AtRPN1b. Mutation of AtRPN1a causes cyclin B1 accumulation, cell division defects and embryo lethality, but the mutant rpn1b does not show an abnormal phenotype.54 The two isoforms RPN5a and RPN5b have redundant and unique functions during gametogenesis and sporophytosis, which is similar to the two isoforms of RPN1.55 Mutation of RPN10 causes the accumulation of ABI5 protein and shows hypersensitivity to ABA.28,56 The mutant rpn12a is less sensitive to exogenous cytokinin.57
Of the six AAA-ATPase subunits in Arabidopsis, only the RPT2 and RPT5 subunits were reported. HLR gene encodes the AtRPT2a subunit, which is necessary for stem cell maintenance in both shoot and root apical meristem.58 Mutation of AtRPT2a results in increased cell expansion and causes large organs, including leaves, stems, seeds and embryos.59 AtRPT5 subunit is encoded by two genes, AtRPT5a and AtRPT5b. AtRPT5a and AtRPT5b have a redundant function in an accession-dependent manner.60 OsRPT4, a subunit of 26S proteasome was screened by yeast two-hybridization using OsRAA1 as bait and their interaction was confirmed in vitro and in vivo assays. OsRAA1 also contains the D-box motif (RGSLDLISL), the common motif for APC recognition, and can be degraded by the ubiquitin-proteasome pathway in both rice and yeast.27
The ubiquitin-proteasome pathway is important for cell cycle regulation during plant development, such as in hormone signaling, embryogenesis and senescence.61 In Arabidopsis, the APC subunits APC2, APC3/HBT, APC/NOMEGA, HBT/CDC27 have been characterized,31,32,34,61 which indicate a conserved role for Arabidopsis APC/C in controlling mitotic progression and cell differentiation during the entire life cycle. Our studies of rice showed that OsRAA1 is degraded by the ubiquitin-proteasome system. First, RAA1-overexpressed rice showed more cells in metaphase than did wild-type rice, and the activity of 26S proteasome inhibited by MG132 and MG115 resulted in more metaphase cells than in untreated rice. Therefore, overexpression of OsRAA1 or the inhibition of 26S proteasome caused metaphase arrest. Second, OsRAA1 degradation in the protein mixture of buds was promoted by ubiquitin and ATP but was delayed by the 26S proteasome inhibitors MG132 or MG115. Third, mutation of the two highly conserved amino acids R and L in D-box lead to OsRAA1 can not be degraded by ubiquitination-dependent manner, which suggests that the D-box is essential for OsRAA1 disruption.27 In yeast and human, the “two hands,” RPN10 and RPN13, can grasp the ubiqutinated substrates.3,4 Although the subunits RPN10 and RPN13 are ubiquitin receptors, which subunit in rice can interact with the ubiquitinated substrates is unknown. Our experiments involving yeast two-hybrid hybridization, in vitro co-immunoprecipitation and in vivo BiFC observation provided clues for the interaction of OsRAA1 and OsRPT4. OsRAA1 may be recognized and ubiquitinated by APC specifically and the ubiquitin-OsRAA1 may be recruited to 26S proteasome by the interaction with OsRPT4, one of the subunits in the 19S regulatory particle. Subsequently, OsRAA1 is degraded by 26S proteasome. The disruption of OsRAA1 is necessary for the transition from metaphase to anaphase during rice root growth.
Although we had reported that OsRAA1 degradation by the ubiquitin-26S proteasome pathway regulates rice root development, some gaps exist in the understanding of protein degradation mediated by the pathway in rice. Further study is needed to investigate the functions of APC components in rice, especially how and which APC activator recognizes OsRAA1 for ubiquitination. How ubiquitinated OsRAA1 interacts with OsRPT4 and the role of other protein subunits of 26S proteasome need to be addressed.
OsRAA1 as a small G protein is co-localized with spindle tubulin during mitosis. It can be recognized by APC. The interaction of OsRAA1 and OsRPT4 leads to its degradation by 26S proteasome. OsRAA1 is a negative regulator for primary root growth. In OsRAA1 transgenic rice, the transition from metaphase to anaphase is blocked by the accumulation of OsRAA1 and reduced growth of primary roots (Fig. 2). OsRAA1-knockdown rice showed altered cell division. During root development, OsRAA1 degradation is required for the normal cell cycle, especially the onset of anaphase, which quantitatively controls the balance of cells in metaphase and anaphase.
This work was supported by the National Nature Science Foundation of China (30670197) and Innovative Program of CAS (KSCX2-YW-N-041).
Previously published online: www.landesbioscience.com/journals/psb/article/10661