Eukaryotic mRNA decay is initiated by shortening of the poly (A) tail; however, neither the molecular mechanisms underlying deadenylation nor its regulation is well understood. The human CCR4–NOT complex is a major cytoplasmic deadenylase consisting of a combination of at least nine subunits, four of which have deadenylase activity. The roles of the other subunits remain obscure. Here, we show that CNOT2 depletion by siRNA induces apoptosis. We also show that CNOT2 depletion destabilizes the complex, resulting in the formation of a complex smaller than that formed in control siRNA-treated cells. The deadenylase activity of the CNOT6L subunit-containing complex prepared from CNOT2-depleted cells was less than that from control cells. Intriguingly, the formation of P-bodies, where mRNA degradation supposedly takes place, was largely suppressed in CNOT2-depleted cells. Furthermore, CNOT2 depletion enhanced CHOP mRNA levels, suggesting that endoplasmic reticulum (ER) stress was occurring, which causes apoptosis in a caspase-dependent manner. These results suggest that CNOT2 is important for controlling cell viability through the maintenance of the structural integrity and enzymatic activity of the CCR4–NOT complex.
Eukaryotic mRNAs have special structures at their 5′ and 3′ ends: a 7-methyl guanosine cap is added to the 5′ end and a poly (A) tail is added to the 3′ end. These structures recruit proteins such as translation initiation factors and poly (A) binding proteins (PABPs), and the interaction of these proteins leads to circularization of the mRNA and facilitates its translation and stabilization. Therefore, the poly (A) tail is thought to regulate mRNA decay and gene expression. Indeed, the decay of most eukaryotic mRNAs follows poly (A) tail shortening (Garneau et al. 2007). Recent studies have shown that specific cytoplasmic RNA granules, known as processing bodies (P-bodies), contain mRNA decay-related proteins such as DCP1/2 decapping enzymes, decapping enhancers, XRN1, argonaute proteins and a helicase family member (RCK/p54). However, P-bodies do not contain either ribosomes or PABPs, which are required for translation (Sheth & Parker 2003; Kedersha et al. 2005; Liu et al. 2005). In addition, the P-body appears to be a dynamic structure formed in the presence of deadenylated mRNAs (Teixeira & Parker 2007; Zheng et al. 2008). These findings suggest that decapping and degradation of mRNAs occur in P-bodies and that the number and size of P-bodies could be a marker of the degree of mRNA decay that is occurring in a cell (Cougot et al. 2004; Eulalio et al. 2007).
The mammalian orthologs of yeast Not1p-3p, Ccr4p, Caf1p and Caf40p are CNOT1-CNOT3, CNOT6/6L, CNOT7/8 and CNOT9, respectively (Albert et al. 2000). The Not4p ortholog CNOT4 is an ubiquitin E3 ligase possessing a RING finger domain (Albert et al. 2002) and is not associated with the complex (Lau et al. 2009). Ccr4p and Caf1p duplicated into CNOT6/6L and CNOT7/8, respectively. There is no human homologue of Caf130p; instead, humans have CNOT10. The functions of the individual subunits are fundamentally conserved, each playing a role in the regulation of gene expression (Collart 2003; Denis & Chen 2003; Collart & Timmers 2004; Morita et al. 2007). However, the precise biological roles of each CNOT subunit remain to be elucidated, and in particular, the function of CNOT2 is poorly understood. Yeast Not2p contributes to the stability of the CCR4–NOT complex (Bai et al. 1999; Russell et al. 2002) and is critical for yeast survival (Russell et al. 2002). In C. elegans and D. melanogaster, Not2 is essential for embryonic development (Frolov et al. 1998; Sonnichsen et al. 2005). Here, we addressed the function of mammalian CNOT2 by depleting it from cultured human cells using short interfering RNAs (siRNAs). We found that CNOT2 depletion destabilized the CCR4–NOT complex and impaired its deadenylase activity. Furthermore, CNOT2 depletion caused apoptosis in a caspase-dependent manner. Thus, in addition to the repression of promoter activity (Zwartjes et al. 2004), human CNOT2 is important for maintaining the deadenylase activity and structural integrity of the CCR4–NOT complex, thereby affecting cell viability.
CNOT2 depletion disrupts the structure of the CCR4–NOT complex
To examine the role of CNOT2 in the organization of the CCR4–NOT complex, we repressed the expression of CNOT2 in HeLa cells by transfecting them with CNOT2 siRNA. The expression of CNOT2 was greatly reduced in the cells treated with CNOT2 siRNA when compared with those treated with control siRNA (Fig. 1A). We also examined the expression levels of the other CCR4–NOT subunits by immunoblotting and found that CNOT2 depletion greatly affected the intracellular levels of CNOT3, whereas the levels of other subunits, such as CNOT6L and CNOT7, were virtually unaffected (Fig. 1A). The levels of CNOT1 and CNOT9 decreased modestly upon CNOT2 depletion. These results suggest that CNOT2 plays a part in maintaining the structural integrity of the CCR4–NOT complex.
Next, we performed gel filtration chromatography using two tandem-connected Superose 6 columns. With samples from control siRNA-treated cells, we found most of the CCR4–NOT subunits in 2.0-MDa fractions (Fig. 1B). The CNOT2 protein migrated predominantly to the 2.0-MDa fraction, whereas other core subunits (CNOT1, CNOT3, CNOT6L, CNOT7 and CNOT9) were found in both the 2.0-MDa fraction and smaller fractions ranging from 0.5 to 1.5 MDa. When the cells were treated with CNOT2 siRNA, the 2.0-MDa form of the CCR4–NOT complex was reduced and most subunits (CNOT1, CNOT6L, CNOT7 and CNOT9) except CNOT3 were found around the ∼1.2-MDa fraction, although some remained in the 2.0-MDa fraction (Fig. 1C,D). Although CNOT3 expression was markedly reduced by CNOT2 depletion (Fig. 1A), it stayed in the 2.0-MDa fraction with a minor population migrating around the 0.5-MDa fraction (Fig. 1C,D). These results suggest that CNOT2 depletion destabilized the CCR4–NOT complex and apparently divided the complex into two parts: the larger part (∼1.2 MDa) that contained CNOT1, CNOT6L, CNOT7 and CNOT9 and the smaller one (∼0.5 MDa) that contained CNOT3. Consistently, although CNOT3 efficiently co-immunoprecipitated other subunits of the CCR4–NOT complex from lysates of control HeLa cells, it only co-immunoprecipitated very low amounts of other subunits of the CCR4–NOT complex from lysates of CNOT2-depleted HeLa cells (Fig. 1A). Furthermore, in vitro GST pull-down assay using purified CNOT3 and GST-CNOT2 showed that CNOT2 interacted with CNOT3 directly (Fig. 1E). Taken together, CNOT2 is important for forming 2.0-MDa authentic CCR4–NOT complex, anchoring two parts (∼1.2-MDa part and ∼0.5-MDa part) of the complex through CNOT3 (Fig. 1F). Our results also showed that CNOT3 could form the 2.0-MDa complex in a CNOT2-independent manner. The mechanism of CNOT2-independent association of CNOT3 with the 2.0-MDa complex is not known (discussed below).
CNOT2 depletion decreases the deadenylase activity of the CCR4–NOT complex in vitro
To determine whether disruption of the CCR4–NOT complex by CNOT2 depletion affects its deadenylase activity, we performed in vitro deadenylase assays. We introduced CNOT2 siRNA into HEK293T cells and then transfected them with an EGFP-CNOT6L expression plasmid. Twenty-four hours after transfection, EGFP-CNOT6L and its associated proteins, including other subunits of the endogenous CCR4–NOT complex (Lau et al. 2009), were purified from the cell lysates using anti-GFP-antibodies. Over-expression of EGFP-CNOT6L did not significantly alter the level of CNOT2 expression, and CNOT2 depletion did not affect the exogenous expression of EGFP-CNOT6L (Fig. 2A). The anti-GFP immunoprecipitates were then incubated with a poly (A) RNA substrate, and the reaction products were analyzed on a denaturing gel. The results show that EGFP-CNOT6L and its associated proteins from control siRNA-treated cells cleaved the poly (A) RNA substrate efficiently (Fig. 2B, lane 3) (Morita et al. 2007; Miyasaka et al. 2008) In contrast, EGFP-CNOT6L and its associated proteins from CNOT2 siRNA-treated cells only weakly deadenylated the poly (A) substrate (Fig. 2B, lane 4). These results suggest that the in vitro deadenylase activity of the CCR4–NOT complex is impaired by CNOT2 depletion.
CNOT2 contributes to the formation of P-bodies
Although CNOT2 is a component of the CCR4–NOT cytoplasmic deadenylase complex, previous reports have shown that CNOT2 is involved in transcriptional regulation in the nucleus (Winkler et al. 2006). Therefore, we first assessed whether the CNOT2 protein is present in the cytoplasm as well as in the nucleus by analyzing the subcellular localization of exogenously expressed mCherry-CNOT2 in HeLa cells. Immunofluorescence analysis showed that mCherry-CNOT2 was largely present in the cytoplasm (Fig. 3A), suggesting that the transcriptional role of CNOT2 is not exclusive. We then asked whether CNOT2 depletion affects the deadenylase activity of the CCR4–NOT complex in vivo. As P-bodies are cytoplasmic RNA granules where mRNA decay is thought to take place, their number in the cells is considered to parallel the degree of mRNA decay (Eulalio et al. 2007). Therefore, we asked whether the number of P-bodies changed as a result of CNOT2 depletion. In control siRNA-treated cells, approximately ten P-bodies, on average, were scattered in the cytoplasm (Fig. 3B,C). In contrast, the number of cytoplasmic P-bodies was significantly reduced in CNOT2-depleted cells (Fig. 3B,C). This raises the possibility that CNOT2 is one of the structural components of the P-body. Co-immunofluorescence studies, however, showed little colocalization of mCherry-CNOT2 with P-bodies visualized by AGO2 immunostaining (Fig. 3D), indicating that CNOT2 is not a structural component of the P-body. Given that the expression level of AGO2 was hardly affected by CNOT2 depletion (Fig. 3E), reduction in the number of P-bodies caused by CNOT2 depletion was because of down-regulation of the activity of the CCR4–NOT complex but not to the alteration in the AGO2 level.
CNOT2 depletion leads to ER stress-mediated cell death
To assess the effect of CNOT2 depletion on cell growth, we analyzed CNOT2-depleted cells with flow cytometry and found that a significant population of CNOT2-depleted cells was present in the sub-G1 fraction. The population of CNOT2-depleted cells in S- and G2/M-phase fractions did not change, compared with the control cells (S-phase fraction: Control siRNA 27.4 ± 4.2%, CNOT2 siRNA #1 21.9 ± 2.6%, CNOT2 siRNA #2 27.9 ± 3.4%. G2/M-phase fraction: Control siRNA 22.1 ± 4.2%, CNOT2 siRNA #1 22.6 ± 0.4%, CNOT2 siRNA #2 28.8 ± 3.0%) (Fig. 4A). These results suggest that CNOT2 depletion induces cell death. Indeed, cleavage of poly ADP ribose polymerase (PARP) and activation of caspase-3 were detected in CNOT2 siRNA-treated cells (Fig. 4B, left panel). Furthermore, administration of the pan-caspase inhibitor z-VAD-fmk effectively suppressed CNOT2 siRNA-induced cleavage of PARP (Fig. 4B, right panel), suggesting that CNOT2 depletion induces apoptotic cell death in a caspase-dependent manner.
Next, we addressed the underlying mechanism of the apoptosis induced by CNOT2 depletion. Because cytoplasmic mRNAs cycle between polysomes and P-bodies (Brengues et al. 2005; Balagopal & Parker 2009), the lack of P-body formation represents poor mRNA deadenylation, possibly leading to an accumulation of mRNAs and, eventually, to an increased translation. If accumulated cytoplasmic mRNAs caused increased translation in cells, CNOT2 depletion could lead to the overproduction of unfolded or misfolded proteins, thereby inducing ER stress and cell death (Kaufman 2002). To address this possibility, we first measured the amount of total proteins in CNOT2-depleted cells by bicinchoninic acid assay (Fig. 5A) and CBB staining (Fig. 5B) and found that CNOT2-depleted cells had a significant increase in protein content per cell compared with control cells. These results suggest that protein overproduction certainly occurred by CNOT2 depletion. We then examined whether the translation inhibitor cycloheximide (CHX) would abrogate apoptotic cell death induced by CNOT2 depletion. CHX is known to be inhibitory to ER stress-mediated apoptosis induced by proteasome inhibitor, bortezomib, as well as to apoptosis induced by TGF-β and serum deprivation (Sanchez et al. 1997; Nawrocki et al. 2005; Bai & Cederbaum 2006). As shown in Fig. 5C, CHX inhibited the cleavage of PARP induced by CNOT2 depletion, suggesting that overproduction of protein products and accumulation of unfolded/misfolded proteins caused the apoptosis in CNOT2-depleted cells.
ER stress-mediated apoptotic cell death is associated with the cleavage of caspase-4, an ortholog of murine caspase-12 (Hitomi et al. 2004). Interestingly, immunoblot analysis showed that caspase-4 cleavage occurred in CNOT2-depleted cells (Fig. 5D). As the transcription of CHOP/GADD153 mRNA occurs only after the cells were exposed to ER stress (Brewer et al. 1997; Zinszner et al. 1998), we analyzed its level by quantitative RT-PCR. The results showed that the amount of CHOP mRNA in CNOT2-depleted cells was approximately three times more than that in control cells (Fig. 5E). These results suggest that CNOT2 depletion induces ER stress and mediates cell death.
Biological importance of CNOT2 has been reported in several model animals. In yeast, Not2p is critically important for viability; Not2p-mutated yeast cells can grow, but the rate of their growth is slower than that of wild-type cells (Collart & Struhl 1994). In the present study, we detected apoptotic cell death upon treatment with CNOT2 siRNA. Apoptosis may be occurring in NOT2-depleted C. elegans and D. melanogaster during early embryonic stages (Frolov et al. 1998; Sonnichsen et al. 2005). As for the molecular function of CNOT2, accumulating results suggest that it plays a role in the regulation of deadenylation as a component of the CCR4–NOT deadenylase complex and/or in transcriptional regulation. Interestingly, we found that simultaneous depletion of all deadenylase subunits also affects cell viability to the same degree of CNOT2 depletion (K. Ito, unpublished observation), suggesting that the deadenylase activity of the CCR4–NOT complex is important for the regulation of cell viability and the attenuation of its deadenylase activity caused by CNOT2 depletion would be responsible for the aberrant apoptosis.
Not2p and Not3p contribute to the formation of the 2.0-MDa CCR4–NOT complex. Consistent with this, we showed here that the formation of the 2.0-MDa CCR4–NOT complex is significantly suppressed by CNOT2 depletion. Moreover, we showed that CNOT2 depletion apparently divided the complex into two parts. Thus, CNOT2 seems to be required for incorporation of the parts including CNOT3 into the CCR4–NOT complex, resulting in generation of the 2.0-MDa CCR4–NOT complex. Association of CNOT3 with the complex was mostly, but not all, abrogated in the absence of CNOT2 (Fig. 1A–D). A significant portion of CNOT3 was detached from the 2.0-MDa form of the complex and ran faster through the gel filtration in the absence of CNOT2 (Fig. 1C). In the immunoprecipitation experiment, only residual association was detected between CNOT3 and CNOT6L, CNOT7 and CNOT9 (Fig. 1A). This discrepancy of the degree of detachment from the complex between the gel filtration and immunoprecipitation may be attributed to different buffer conditions. The precise mechanism of association of CNOT3 with the complex in the absence of CNOT2 remains to be understood. Because CNOT3 possesses the short sequence motif with similarity to CNOT2 (Zwartjes et al. 2004), it is possible that CNOT3 interacts with the complex through its CNOT2-homology domain. Thus, CNOT3 would be incorporated into the complex in two ways: a direct but weak interaction with the complex and an indirect but strong interaction mediated by CNOT2 as illustrated in Fig. 1F.
Our results suggest that the disruption of the CCR4–NOT complex induced by CNOT2 depletion is accompanied by a loss of deadenylase activity. Consistent with this, previous reports showed that knockdown of Regena/NOT2 inhibits mRNA decay in Drosophila S2 cells (Temme et al. 2004). How CNOT2 contributes to the regulation of the deadenylase activity of the CCR4–NOT complex remains obscure. Note that depletion of CNOT1, a scaffold subunit of the complex, also suppresses P-body formation and mRNA decay through the deterioration of the CCR4–NOT complex (K. Ito, unpublished observation). Thus, the structural integrity of the complex is apparently important for the full exhibition of the deadenylase activity. We also found that CNOT3 could enhance the deadenylase activity of the CCR4–NOT complex in vitro (M. Morita, unpublished observation). Several reports suggested that the CCR4–NOT complex interacts with RNA binding proteins, such as NANOS2, to access and degrade its target mRNAs in vivo (Suzuki et al. 2010). Such RNA binding proteins may interact with the complex to form the ∼2.0-MDa form through CNOT2 as is CNOT3 is incorporated into the complex, and CNOT2 might thus be a harbor for proteins that regulate the deadenylase activity of the CCR4–NOT complex.
We provided evidence that CNOT2 depletion causes overproduction of proteins due most likely to enhanced translation from stabilized mRNAs, which in turn would lead to the generation of unfolded/misfolded proteins. Although the presence of the unfolded/misfolded protein was not showed directly, our molecular biological studies provided strong evidence of their accumulation in CNOT2-depleted cells (Fig. 5C–E). Excess of unfolded or misfolded proteins in CNOT2-depleted cells could trigger ER stress, which in turn would up-regulate transcription of mRNA of the transcription factor CHOP (Wang et al. 1996; Zinszner et al. 1998). Alternatively, as CNOT2 depletion ameliorates the deadenylase activity of the CCR4–NOT complex, the CHOP mRNA may be poorly deadenylated and stabilized in the CNOT2-depleted cells. The CHOP mRNA is barely expressed under normal conditions, and its increased expression could amplify the pro-apoptotic signal by altering the balance between Bcl-2 and Bax (McCullough et al. 2001). Therefore, we propose that apoptosis induced by CNOT2 depletion is mediated by up-regulation of the CHOP mRNA. It should be also noted that enhanced translation increases reactive oxygen species (ROS) that is generated from oxidative protein folding in ER (Sevier & Kaiser 2008). ROS is involved in various biological processes and its increased production would induce apoptosis (Ichijo 1999). Further molecular and biochemical analyses of CNOT2-depleted cells as well as Cnot2 knockout mice would help clarify the precise mechanism of the apoptosis.
Antibodies and plasmids
Rabbit anti-CNOT1, CNOT3, CNOT6L, CNOT7 and CNOT9 polyclonal antibodies were generated by immunizing rabbits (Morita et al. 2007). Anti-CNOT2 monoclonal antibody was obtained from Biomatrix. Anti-PARP and anti-α-tubulin antibodies were from ZYMED and Sigma, respectively. Anti-GFP antibody and anti-caspase4 antibody were from MBL. Anti-human AGO2 monoclonal antibody (4B1) was from WAKO. The expression plasmid for CNOT6L (pEGFP-CNOT6L) was described previously (Morita et al. 2007; Miyasaka et al. 2008). The human CNOT2 cDNA was obtained by reverse transcription (RT)-PCR of total RNA prepared from HeLa cells. A full-length cDNA encoding CNOT2 was inserted into the pmCherry-C1 expression plasmid (Clontech).
Cell culture and transfection
HeLa cells and HEK293T cells were cultured in Dulbecco’s modified Eagle’s medium (Nissui) containing 10% fetal bovine serum (HyClone). RNA oligonucleotides targeting the following cDNA sequences were synthesized (Sigma Aldrich) and annealed using standard protocols: 5′- UUCUCCGAACGUGUCACGUTT -3′ for control siRNA, 5′- CUAGCAGGACAAAUAGCAUGA -3′ for CNOT2 siRNA #1 and 5′-CCACGUCACGCCAACAACAGG-3′ for CNOT2 siRNA #2. These siRNAs were transfected into HeLa cells and HEK293T cells using Oligofectamine™ RNAi MAX (Invitrogen). Seventy-two hours after transfection, cells were harvested and used for further experiments. Plasmid DNAs were transfected into HeLa cells using the FuGENE® 6 Transfection Reagent (Roche). HEK293T cells were transfected with plasmid DNA by the calcium-phosphate transfection method.
Immunoprecipitation and immunoblotting
Cells were lysed with 1% NP-40 TNE buffer [50 mm Tris–HCl (pH 7.5), 1 mm EDTA, 150 mm NaCl, 1 mm NaF] on ice for 15 min and clarified by centrifuging. For immunoprecipitation, the lysates were incubated with the appropriate antibodies for an hour at 4 °C. Then, 20 μL of Protein G-Sepharose (50% slurry) (GE healthcare) was added to the samples and the mixtures were incubated for another 2 h at 4 °C. The beads were washed with TNE buffer four times and resuspended with SDS-PAGE loading buffer. The samples were separated by SDS-PAGE and transferred to Immobilon™ Transfer Membrane (Millipore). Immunoblotting was performed as described previously (Morita et al. 2007).
HEK293T cells were transfected with CNOT2 siRNAs. Twenty-four hours after siRNA transfection, cells were transfected with EGFP-CNOT6L expression plasmids. After another 24 h, lysates were prepared from the cells and immunoprecipitated with an anti-GFP antibody. The precipitates were incubated at 37 °C for 180 min with synthesized RNA substrate (5′-UCUAAAUAAAAAAAAAAAAAAAAAAAA-3′; final concentration, 0.1 μm) labeled with fluorescein isothiocyanate at the 5′-end. The reaction products were fractionated on a 7 m urea–25% polyacrylamide denaturing gel. The gel was analyzed with an FLA-5000 Fluorescence Imager (Fujifilm).
HeLa cells were grown on glass cover slips and fixed with 2% paraformaldehyde/PBS at room temperature for 10 min. Immunofluorescence staining was performed as described previously (Morita et al. 2007). Photomicrographs were obtained using a Fluoview FV1000 confocal laser scanning biological microscope (Olympus).
Total RNA was extracted from cells using ISOGEN (Nippon Gene) and subjected to reverse transcription using Superscript™ III reverse transcriptase and oligo dT12-18 primers (Invitrogen). Quantitative real-time RT-PCR was then performed using the ABI Prism 7900HT sequence detection system (Applied Biosystems). The primer sets used in this study are as described previously (Lin et al. 2007).
Gel filtration chromatography
Protein lysates were applied on two Superose™ 6 10/300 GL columns connected in tandem, and the columns were eluted with running buffer that consisted of 40 mm Tris–HCl (pH 7.4), 150 mm NaCl, 5 mm EDTA and 0.1% NP-40. The flow rate was 0.2 mL/min, and 0.3 mL was collected per fraction. Molecular weights for each fraction were calculated based on the elution volumes of blue dextran, thyroglobin, ferritin and aldolase.
Protein expression and purification
Full-length cDNA of human CNOT2 and CNOT3 and DNA fragment containing sequence for GST from the pGEX-6P-1 vector were cloned into the baculovirus transfer vector pFastBac 1 (Invitrogen), and the recombinant baculoviruses encoding GST-CNOT2 and GST-CNOT3 were obtained using the Bac-to-Bac™ system (Invitrogen) according to the manufacturer’s instructions. Sf9 cells infected with the recombinant baculoviruses were harvested after 48 h by centrifugation and lysed with 1% NP-40 TNE buffer. The soluble GST-CNOT2 and GST-CNOT3 proteins were purified with Glutathione Sepharose™ 4B (Amersham Bioscience). GST-CNOT3 was cleaved with PreScission™ Protease at 4 °C overnight.
GST ‘pull-down’ experiments in vitro
Recombinant CNOT3 protein was incubated with GST- or GST-CNOT2-glutathione-Sepharose beads overnight at 4 °C. After washing three times with the TNE buffer, the proteins retained on the beads were analyzed on an SDS–PAGE gel.
Measurement of protein content
Forty-eight hours after siRNA transfection, cells were suspended in PBS and the number of cells was counted with the hemocytometer. Cells were then lysed and soluble protein content was measured using the BCA Protein Assay Reagent (Thermo scientific).
We thank Daisuke Ejima (Ajinomoto Co. Inc.) for his kind assistance with gel filtration chromatography analysis. This work was supported in part by research fellowship of the Japan Society for the Promotion of Science (JSPS) for Young Scientists, the Global COE program (Integrative life science based on the study of biosignaling mechanisms) and grants-in-aid from JSPS and from the Ministry of Education, Cultures, Sports, Science and Technology (MEXT), Japan.