|Home | About | Journals | Submit | Contact Us | Français|
ABCE1 is a highly conserved protein universally present in eukaryotes and archaea, which is crucial for the viability of different organisms. First identified as RNase L inhibitor, ABCE1 is currently recognized as an essential translation factor involved in several stages of eukaryotic translation and ribosome biogenesis. The nature of vital functions of ABCE1, however, remains unexplained. Here, we study the role of ABCE1 in human cell proliferation and its possible connection to translation. We show that ABCE1 depletion by siRNA results in a decreased rate of cell growth due to accumulation of cells in S phase, which is accompanied by inefficient DNA synthesis and reduced histone mRNA and protein levels. We infer that in addition to the role in general translation, ABCE1 is involved in histone biosynthesis and DNA replication and therefore is essential for normal S phase progression. In addition, we analyze whether ABCE1 is implicated in transcript-specific translation via its association with the eIF3 complex subunits known to control the synthesis of cell proliferation-related proteins. The expression levels of a few such targets regulated by eIF3A, however, were not consistently affected by ABCE1 depletion.
ABCE1 (also named RNase L inhibitor, Rli1p in yeast and Pixie in Drosophila) is one of the most conserved proteins in evolution, which is universally present in eukaryotes and archaea. It belongs to the family of ATP-binding cassette (ABC) transporter proteins, which carry out a broad range of functions mainly in translocation of substrates across membranes, but are also involved in translation, DNA repair and chromosome maintenance.1,2 In contrast to most ABC proteins, ABCE1 lacks the transmembrane domains and therefore is unlikely to participate in membrane transport.3 ABC proteins typically contain twin nucleotide-binding domains (NBDs) that act as molecular engines coupling ATP hydrolysis to conformational changes in their substrate-binding sites. The two NBDs of ABCE1 are arranged in a head-to-tail orientation through the hinge domain.4 Unlike all other ABC enzymes, ABCE1 contains the N-terminal domain with two iron-sulfur (FeS) clusters.5
In all studied organisms, loss in ABCE1 function appears to be lethal or causing severe defects in growth and development.6-10 Human ABCE1 was first characterized as an inhibitor of RNase L, a latent endoribonuclease functioning on the interferon-dependent antiviral 2–5A pathway.11 Besides antiviral activity, this pathway plays a more general role in regulation of cellular RNA expression and turnover.12,13 In contrast to ABCE1, RNase L is present only in vertebrates, implying that ABCE1 must have essential functions not related to the 2–5A pathway.
One fundamental function of ABCE1 could be its involvement in ribosome biogenesis, which is coupled to FeS cluster maturation in yeast.14,15 A number of recent studies have established a role of ABCE1 in several steps of eukaryotic translation. ABCE1 has been shown to associate with 40S ribosomal subunits and initiation factors eIF2, eIF3, and eIF5, suggesting its role in pre-initiation complex assembly.15-18 Consistently, ABCE1 depletion results in reduced polysome content and global translation in yeast, mammalian and Drosophila cells.10,14,16-18 ABCE1 is also involved in translation termination through interaction with release factors eRF1 and eRF3, occupying the GTPase center.19,20 Interestingly, its function in termination could be shared with RNase L that might balance mRNA turnover and translation.21 Besides promoting translation termination, ABCE1 can act as an efficient ribosome recycling factor catalyzing post-termination complex dissociation at various conditions.22 In addition to canonical termination, ABCE1 dissociates stalled and vacant ribosomes acting in quality control mechanisms during mRNA translation and ribosome biogenesis, and in regulation of available ribosome pool.23-25 Moreover, ABCE1 recycling activity may control non-canonical 3′-UTR translation on stalled ribosomes during stress.26 The function of ABCE1 in ribosome recycling is conserved in archaea, in contrast to the initiation stage lacking the eukaryotic homologues of eIF3 and eIF5.27
Thus, ABCE1 role in translation could be considered as a link between termination, recycling and initiation.28,29 Notably, ABCE1 may directly associate with yeast Hcr1 (eIF3J in higher eukaryotes) subunit of eIF3, another multifunctional translation factor involved in the same steps of translation and in rRNA processing.15,30-32 Furthermore, the diverse roles of ABCE1 include HIV capsid assembly and endogenous suppression of RNA interference recently demonstrated in several organisms.33,34
Despite a substantial progress in understanding the various ABCE1 functions, it remains unexplained which of them are critical for cell viability and growth. Due to its central biological role, the involvement of ABCE1 in pathogenesis could be expected, although this knowledge is currently rather limited. Upregulated ABCE1 expression is increasingly found in association with cancer, whereas inhibition of ABCE1 can efficiently suppress tumor cell proliferation.35-39 Further research on ABCE1 function and regulation may therefore contribute to the development of efficient anticancer therapy strategies.
An indirect relevance of ABCE1 to tumorigenesis may concern its function as an inhibitor of RNase L, which is implicated in hereditary prostate cancer.40,41 As demonstrated by several studies, modulation of ABCE1 expression levels appears to correlate with the respective changes in RNase L activity.12,42,43 Furthermore, a decreased gene expression of ABCE1 has been suggested to account for an increased activity of RNase L in patients with chronic fatigue syndrome, another pathology associated with the 2–5A pathway.44
In the current study, we address the underlying mechanism of ABCE1 involvement in the regulation of proliferation of cultured human cells and its possible connection to the fundamental role of ABCE1 in translation. We speculate that besides its function in basal translation, ABCE1 could be required for histone biosynthesis and DNA replication hence providing normal S phase progression. In addition, we analyze whether ABCE1 could be implicated in transcript-specific translation via its association with eIF3 complex subunits known to control the synthesis of several cell proliferation-related proteins.
The role of ABCE1 in cell proliferation was studied using cultured human cells (HEK293 and HeLa) with downregulated ABCE1 expression. For that purpose, two ABCE1-specific siRNA constructs (ABCE1-1 and ABCE1-2) as well as a scrambled control (Scr) siRNA were applied. The relative growth of cells was estimated by the MTT assay on days 3 and 6 after transfection, and the efficiency of silencing was monitored by the method of qPCR.
In both cell lines, ABCE1 silencing efficiency was reaching the values of up to 90% already on day 3 after transfection, which however produced only a minor effect on the growth phenotype (Fig. 1A and B). In contrast, the effect on cell proliferation was substantial (50–70% of the control value) at the prolonged culturing of the transfected cells for 6 day (with an intermediate passage on day 3). These results confirm the previous reports on ABCE1 depletion inhibiting HEK293 and cancer cell proliferation, although a more drastic effect has been shown before for the HEK293 cells.17,38,39 In our study, the effect of ABCE1 depletion on cell growth was dependent on cell confluence, usually producing stronger phenotype at low densities, up to a failure to proliferate (data not shown). Therefore, cells at intermediate confluence were used in the growth assay. Morphologically, ABCE1-depleted cells were abnormally looking (Fig. 1C): smaller in size, round, less dense, with some cells detached, in particular when plated at low density. In general, ABCE1-1 siRNA produced more efficient silencing (reaching the values of 70–90% at the protein level) and was therefore chosen for further experiments.
To study the observed effect on cell proliferation in more detail, we next analyzed cell cycle profile in ABCE1-depleted HEK293 cells by flow cytometry. ABCE1 depletion was associated with a pronounced defect in cell cycle profile distribution, producing a 1.7-fold increase of cells in S phase and a 0.6-fold decrease in G0/G1 phase (Fig. 2A). The effect was observed already on day 4 (the next day after the passage of cells) and persisted until day 6 after transfection. A similar change in cell cycle profile was detected also in HeLa cells (Fig. S1A). Previously, a comparable effect of ABCE1 silencing on HeLa cell cycle phenotype was obtained using a genome-scale RNAi screening analysis, where ABCE1 was found among the genes essential for cell division.45
In order to determine the affected step in cell cycle progression, HEK293 cells with depleted ABCE1 levels were synchronized at the G1/S boundary by a double thymidine block and released into S phase. This analysis revealed that the G1/S transition was not affected, whereas the delay in cell cycle progression occurred at mid-S phase, between 2 and 4 h after release (Fig. 2B).
RNase L, the binding substrate of ABCE1, is known to regulate the stability of some mRNAs involved in cell growth control, such as HuR and TTP that subsequently act to modulate the expression of e.g. p21 governing G1/S transition.46-48 To make sure that the effect on cell cycle progression observed with depleted ABCE1 does not depend on any mechanism promoted by RNase L activation (either as a result of ABCE1 inhibition or non-specific off-target effect of the applied siRNAs), RNase L expression was downregulated by siRNA either alone or in combination with ABCE1 (Fig. S2A). The transfected cells were synchronized and released into S phase. Whereas RNase L depletion increased the proportion of cells entering S phase, possibly as a consequence of p21 deregulation,47 it had no subsequent effect on S phase progression (Fig. S2B). In contrast, the combination of RNase L and ABCE1 silencing resulted in S phase delay as observed in the case of ABCE1 depletion alone. We therefore conclude that the observed S phase progression defect in ABCE1-depleted cells is a genuine effect of ABCE1 downregulation, which is independent of the RNase L pathway.
In summary, the prolonged depletion of ABCE1 is associated with abnormal morphology and impaired growth of HEK293 and HeLa cells. This phenotype is accompanied by a defect in cell cycle profile distribution due to a delay in S phase progression.
As shown above, ABCE1 depletion resulted in a slow progression of cells through S phase without affecting S phase entry. We supposed that this effect might be caused by intra-S phase checkpoint activation, a mechanism triggered by aberrant events during DNA replication that results in a transient decrease in the rate of DNA synthesis.49,50 We therefore decided to analyze whether ABCE1 is essential for DNA replication. To study DNA de novo synthesis, the transfected HEK293 cells were synchronized by a double thymidine block and released into S phase on day 6 after transfection. The DNA was labeled by bromodeoxyuridine (BrdU) and cells were collected for a time period of 1 to 6 h covering S phase. In cells with downregulated ABCE1, the proportion of DNA-synthetizing cells was significantly lower at 2 and 4 h compared to control (Fig. 3A), which correspond to the period when the S phase progression gets delayed. Thus, insufficient DNA synthesis could limit the rate of cell cycle progression in ABCE1-depleted cells, which may explain their accumulation in mid-S phase.
Besides a critical role in replication stress response, the S phase checkpoint pathway is also involved in the control of coupling between DNA replication and histone protein synthesis that must be tightly balanced.51 As shown previously, inhibition of histone synthesis via the depletion of the stem-loop binding protein (SLBP) resulted in a decreased rate of DNA synthesis mediated by the S phase checkpoint activation, which also slowed down S phase progression.52-54 To investigate a possible link between DNA synthesis and histone expression in our study, we analyzed the steady-state levels of two core histone proteins, H2B and H4, which indeed appeared to be clearly reduced (to 55–60% of the control value) in ABCE1-depleted HEK293 cells (Fig. 3B). A strong effect on histone protein levels was observed also in HeLa cells (Fig. S3). Since DNA synthesis inhibition leads to rapid destabilization of histone mRNA,55 we next followed H2B and H4 mRNA levels in synchronized HEK293 cells. In ABCE1-silenced cells, the expression of both histone mRNAs during the entire S phase remained significantly lower than in control cells, and their steady-state levels in asynchronous cells were reduced more than twofold (Fig. 3C).
Our results thus confirm that both DNA synthesis and histone expression are reduced in ABCE1-depleted cells. These interdependent effects could be explained by ABCE1 role in histone biosynthesis as a translation factor, although its involvement in DNA replication might be also hypothesized, since a partial nuclear localization of yeast ABCE1 homolog has been reported.16 To confirm whether ABCE1 presence in the nucleus is conserved also in mammalian cells, HeLa cells were labeled with anti-ABCE1 antibody and nuclei visualized by DNA staining. The confocal microscopy analysis revealed that ABCE1 is present both in the nucleus and cytoplasm (Fig. 3D). Thus, any potential nuclear functions of human ABCE1 cannot be excluded.
Since ABCE1 is recognized as an essential translation factor in several organisms,10,14,17 we considered it plausible that ABCE1 could be implicated in histone synthesis via its general function in translation. We therefore decided to examine the status of total protein synthesis in ABCE1-downregulated cells. The depletion of human ABCE1 has been shown to severely inhibit global translation in HEK293 cells and in vitro.17 Similarly, the depletion of ABCE1 homolog, Rli1p, caused a strong inhibition of translation in yeast cells, but only when grown in glucose-rich medium.14
Translation levels in ABCE1-depleted HEK293 and HeLa cells were studied by [35S]-methionine pulse labeling of the newly synthesized proteins on day 6 after transfection. In both cell lines, ABCE1 depletion inhibited de novo protein synthesis, although the effect in HeLa cells was generally stronger, probably due to the metabolic differences between the two cell-lines and overall more efficient silencing of the ABCE1 protein in HeLa cells (Fig. 4A and B). In average, the remaining levels of total translation in ABCE1-depleted HEK293 and HeLa cells were about 50% and 20% of the control values, respectively.
Notably, the extent of translation inhibition varied between the experiments, probably depending on the exact degree of ABCE1 silencing. This implies that in cultured cells, relatively low threshold of ABCE1 expression could be critical for the maintenance of general protein synthesis. In contrast, the cell cycle defect in ABCE1-depleted cells was more evident and observed even in cells with minimal changes in morphology (data not shown), suggesting a higher vulnerability of proper S phase progression to ABCE1 inhibition.
To find out whether the defect in cell cycle could be a consequence of impaired translation, the studies on translation and cell cycle were performed in parallel in HEK293 cells on day 4 after transfection. By that time, the accumulation of cells in S phase was clearly observed, whereas the levels of maintained translation were still considerable (60 – 80% of the control values) (Fig. 4C). Similarly, the defect in cell cycle preceding any significant effect on translation was observed also in ABCE1-depleted HeLa cells (Fig. S1A and B). Our data suggest that the cell cycle defect in ABCE1-downregulated cells occurs before a dramatic depletion of total protein synthesis, presumably as a result of reduced histone synthesis regulated independently of general translation.
In order to further address the involvement of ABCE1 in S phase progression, we analyzed the endogenous ABCE1 expression pattern during S phase. HEK293 cells were synchronized by a double thymidine block and released into S phase, and the ABCE1 protein levels were followed during the next 8 hours. As a control, we followed the expression levels of SLBP protein known to accumulate during S phase.56 We observed that ABCE1 protein expression levels remained stable throughout S phase (Fig. 5). The results of the whole-genome screening of the periodically expressed genes in HeLa cells also suggest no cell cycle-specific oscillations in ABCE1 mRNA expression pattern.57 ABCE1 expression thus appears to be non-periodic, which could be consistent with its role in general translation.
The assumption that histone synthesis control in ABCE1-depleted cells occurs independently of general translation led us to consider whether ABCE1 could be specifically involved in the synthesis of any regulatory proteins essential for cell proliferation. The translation of such potential targets may imply selective initiation on certain mRNAs, which has been suggested for several eIF3 complex subunits regulating cell proliferation.58-60
Remarkably, yeast and Drosophila ABCE1 homologues have been shown to associate with several eIF3 subunits,16,18 some of which are also predicted as putative interactors of human ABCE1 by the STRING 10 database61 (Fig. S4A). Two subunits, eIF3H and eIF3F, were previously found to associate with human ABCE1 within a global analysis of deubiquitinating interactome.62 For further studies on human ABCE1 interacting partners, we selected the eIF3A and eIF3C subunits, which belong to the functional core of mammalian eIF3 complex and play a critical role in its integrity.63,64 Co-immunoprecipitations with anti-ABCE1 antibody were performed on HEK293 whole cell lysates, and the precipitated proteins were analyzed by Western blot. Both eIF3A and eIF3C were detected at rather low intensity in several experiments (Fig. S4B). In addition, the reverse precipitations with anti-eIF3A or anti-eIF3C antibodies failed to detect ABCE1 with reliable confidence (data not shown). This suggests that ABCE1 interaction with the eIF3A and eIF3C subunits might be indirect or labile under the applied conditions.
ABCE1 homologues in yeast and Drosophila may not be the integral components of eIF3 and are rather proposed to facilitate the assembly of translation pre-initiation complexes.16,18 Furthermore, a maximal depletion of the Drosophila homolog, Pixie, reduced the protein levels of some core eIF3 subunits, suggesting a role in their stabilization.18 Since eIF3A is critical for the assembly of the entire eIF3 complex in human cells,64 we analyzed whether ABCE1 depletion could affect eIF3A abundance and therefore cause eIF3 instability. The abundance of eIF3A and another core subunit, eIF3H, were not changed in HEK293 cells with downregulated ABCE1 (Fig. 6A), implying that at least at the level of depletion reached in our study, ABCE1 may not be crucial for their stability.
Considering the essential role of ABCE1 in the initiation step of translation,16-18 we next intended to analyze whether ABCE1 depletion may affect the expression of proteins specifically regulated by eIF3 subunits. In particular, eIF3A is expressed periodically during S phase and may regulate translation of several proteins important for cell cycle control and DNA replication, such as the cyclin-dependent kinase inhibitor p27, ribonucleotide reductase M2 subunit (RRM2) and replication protein RPA2.65-68 eIF3A has been previously shown to stimulate RRM2 and to inhibit p27 and RPA2 translation, thus acting in reciprocal directions.66-68
In our study, in both HEK293 and HeLa cells with depleted ABCE1, the expression of none of these proteins was consistently altered as if they were co-regulated by ABCE1 (Fig. 6B and C). It should be mentioned that in contrast to the unchanged steady-state levels of RRM2 and RPA2 proteins, p27 expression was moderately increased in some experiments with HEK293 cells, but decreased in HeLa cells. The reason for this discrepancy remains unclear, but could be rather attributed to a complex regulation of p27 translation and turnover.69,70 In fact, the decreased levels of all three studied proteins as well as eIF3A were also observed in some experiments with HeLa cells, which could be explained by a drop in total translation efficiency below a critical level due to more efficient ABCE1 depletion. In summary, our initial results suggest that ABCE1 may not be specifically required for the translation of eIF3A-regulated proteins, although more potential targets must be further analyzed for a comprehensive conclusion.
In our study, we investigated the role of ABCE1 in human cell proliferation and cell cycle progression using RNAi approach. The major phenotype of ABCE1-depleted cells was a significant growth inhibition and abnormal cell cycle profile distribution with a higher proportion of cells in S phase and a decreased proportion in G0/G1 phase (Fig. 1 and and2A).2A). Further analysis on synchronized cells revealed that the defect in cell cycle occurs during S phase progression, but does not affect S phase entry (Fig. 2B). This led us to assume that ABCE1 downregulation may interfere with the essential events within S phase.
To elucidate the role of ABCE1 in S phase progression, we first investigated DNA synthesis in ABCE1-depleted cells and found that its rate was significantly inhibited (Fig. 3A). To approach the mechanism of inefficient DNA synthesis, we considered the fact that DNA replication is tightly coupled to histone protein synthesis and chromatin assembly during S phase, so that the rate of DNA synthesis is closely dependent on new histone supply.71,72 Indeed, as demonstrated by several studies, the transcriptional repression of histone mRNA or depletion of the essential histone mRNA regulators in mammalian cells result in inhibition of DNA synthesis and S phase progression arrest.53,73-75 A similar effect has been shown to be caused by the inactivation of chromatin assembly system, which was also accompanied by DNA damage accumulation.76,77
In order to check whether ABCE1 silencing affects histone biosynthesis, we analyzed the steady-state levels of two core histone proteins, H2B and H4, as well as the expression of their mRNAs during S phase, and those appeared to be significantly reduced (Fig. 3B and C, Fig. S3). This effect could be relevant to all replication-dependent histones, whose stoichiometry must be highly coordinated.78,79 We thus suggest that ABCE1 depletion may inhibit histone biosynthesis that results in a suppression of DNA replication and slow cell cycle progression through S phase. Furthermore, our observation that the defect in cell cycle progression occurs after the entry into S phase (Fig. 2B) could be consistent with the mechanism of coupling of DNA and histone synthesis, which is independent of cyclin E/cdk2 activity controlling G1/S transition.73,80,81
Conceivably, ABCE1 deficiency could also impair proper chromatin assembly further impeding M phase progression, which might account for a reduced number of cells progressing into G0/G1 phase (Fig. 2A). Besides the surmised role in histone synthesis, ABCE1 could be anticipated to influence the processes of DNA replication and/or chromatin assembly via its presence in the nucleus (Fig. 3D). Although no such functions of ABCE1 are currently described, several proteins involved in chromatin assembly and remodeling (e.g., SUPT16H, SMARCA5, CBX3 and CBX5) as well as in DNA replication and repair (e.g., RFC4, MMS19) were recently identified among the putative ABCE1 interactors.34,82
A disruption of SLBP, a central regulator of histone mRNA metabolism, is deleterious for embryonic development of several organisms, resulting in chromosome condensation defect in Drosophila and C. elegans83,84 and DNA replication failure in mouse oocytes.85 Human somatic cells with depleted SLBP are viable, although their growth rate is significantly impaired due to a delayed progression through S phase as a result of inefficient DNA synthesis,52-54 thus producing a similar phenotype as ABCE1 silencing in our study. It is hence tempting to speculate that the involvement in histone synthesis may be one of the vital functions of ABCE1. However, the control of cell proliferation by ABCE1 could be more complex due to its fundamental role in translation as well as implication in the RNase L pathway regulating diverse cellular activities.13,29
Histone gene expression is restricted to S phase and is regulated by several mechanisms at the transcriptional and post-transcriptional levels.79,86 Efficient translation of histone mRNAs lacking polyA tails is maintained by their circularized structure, which is stabilized by multiple protein-protein interactions mediated by SLBP bound to the stem-loop sequence at the 3′ end.87 The mechanism of ABCE1 involvement in histone biosynthesis remains to be established, but presumably could be related to its well-recognized function in translation. Since ABCE1 depletion has been previously shown to inhibit total translation in several organisms,10,16,17 we decided to investigate whether insufficient histone synthesis in ABCE1-downregulated cells could be a consequence of general translation deficiency. Consistently with the previous reports, the total protein synthesis in ABCE1-depleted cells was significantly impaired (Fig. 4A and B), although in our system, the extent of inhibition was variable, probably due to small differences in the experimental conditions and the resulting efficiency of ABCE1 silencing. Remarkably, however, the defect in cell cycle progression was not obviously influenced by these variations.
Non-specific inhibition of translation in mammalian cells can accordingly suppress DNA synthesis as a result of histone protein deficiency.88,89 Therefore, we further compared the translation efficiency in ABCE1-depleted cells with their cell cycle profiles, which would reflect any changes in the expression of histone proteins. Importantly, we observed the defect in cell cycle progression occurring before any significant drop in global translation (Fig. 4C, Fig. S1A and B), implying that the efficiency of histone protein synthesis in ABCE1-depleted cells may not correlate with the status of general translation. Namely, a higher threshold of ABCE1 expression level could be required to maintain sufficient histone synthesis compared to the majority of proteins, provided that this regulation occurs at the translational level. Plausibly, this could be explained by a high demand for massive production of histones during S phase, whereas only a very small pool of free histones is available in mammalian somatic cells due to their rapid incorporation into nucleosomes.90,91 In addition, ABCE1 might be particularly required for the recruitment and proper tethering of translation factors in the mRNP complex, providing stability and efficient translation of histone mRNAs. ABCE1 function in translation termination is also likely to regulate histone mRNA stability, which is controlled by Upf1-mediated decay mechanism coupled to translation and normally triggered when DNA replication is inhibited.86,87 Inefficient termination is proposed to determine histone mRNA degradation92,93 and thus could account for the reduced histone mRNA levels observed in ABCE1-depleted cells (Fig. 3C).
In order to examine if ABCE1 expression correlates with the period of histone synthesis, we followed ABCE1 protein levels in synchronized cells, which appeared to be unchanged during S phase (Fig. 5). This is in contrast to the SLBP expression, which rapidly increases at the translational level just before the accumulation of histone mRNA.56 ABCE1 therefore is unlikely to act as a specific regulator of histone proteins and may rather facilitate their efficient synthesis as speculated above.
At present, there is no evidence of ABCE1 involvement in transcript-specific translation, although ABCE1-mediated ribosome recycling is predicted to greatly influence the efficiency of translation of some regulatory mRNAs94 and to control translation re-initiation events under certain conditions.26,95 To get further insight into the involvement of ABCE1 in cell proliferation, we considered its potential role in the synthesis of proteins essential for cell cycle progression. The translation of certain transcripts is highly dependent on the initiation step, which is determined by both the activity of initiation factors and the 5′-UTR structure of mRNA.96 ABCE1 has been suggested to have an essential function in pre-initiation complex assembly through association with a number of initiation factors.16-18 Among them, the eIF3 complex plays a central role in initiation of general translation, but is also known to regulate synthesis of specific proteins involved in cell growth control as well as replication-dependent histones due to its alternative composition in eukaryotes.58,59,97 A selective translation of cell growth-specific mRNAs may also be responsible for the implication of several eIF3 subunits in oncogenesis.60,98
In our study, we attempted to address a potential role of ABCE1 in eIF3-related translation. First, we tested whether ABCE1 depletion may decrease eIF3 stability, an effect previously suggested for the ABCE1 homolog in Drosophila.18 Two eIF3 core subunits, eIF3A and eIF3H, were chosen, both of which are suggested to be involved in translation of certain mRNAs.58,59,63,99 Furthermore, eIF3A, the largest and conserved eIF3 subunit, may control the assembly and stability of eIF3 as well as the formation of different sub-complexes with additional functions within the mammalian cells.64,100 However, differently from the Drosophila homolog, ABCE1 silencing in our system did not reduce the levels of eIF3A and eIF3H (Fig. 6A), as previously demonstrated also for several initiation factors in yeast.16,18
Then, we analyzed if ABCE1 depletion may affect translation of any eIF3-specific transcripts, focusing on the known eIF3A-regulated targets involved in cell cycle control and DNA replication. Among them are p27, an essential regulator of S phase entry; RRM2, a key enzyme supplying converted deoxyribonucleotides for DNA synthesis; and RPA2 functioning in multiple DNA metabolic pathways.66-68 The steady-state levels of these proteins were also not consistently altered in ABCE1-silenced cells (Fig. 6B), implying that translation of at least these few candidates may not be critically dependent on the presence of ABCE1. Instead, the translation of cell growth-specific transcripts could be primarily determined by eIF3 binding to the 5′-UTR of defined mRNAs.98
Besides the initiation step, both ABCE1 and eIF3 are involved in termination and ribosome recycling, as well as in various translation-linked processes, thus acting as unique multifunctional translation factors.22,31,32* Moreover, the eIF3 complex in S. pombe is proposed to be involved in global regulation of translation due to its presence in several complexes involved in protein synthesis and degradation.101 In addition to linking several stages of translation, ABCE1 is required for rRNA processing and nuclear export, whereas its universal function in recycling could be implicated in ribosome maturation, mRNA and protein quality control pathways.15,23,24,102
In conclusion, we suggest here a novel link between ABCE1 role in translation, DNA synthesis and cell cycle progression via the control of histone protein synthesis. Since ABCE1 levels are particularly sensitive to oxidative stress,103 such regulation could be further connected to cellular metabolism and act to prevent DNA replication damage during the unfavorable conditions. Another level of S phase progression control is mediated by GAPDH, which in addition to its traditional glycolytic function, appears to be involved in histone gene transcriptional regulation in response to cellular redox state.74,104 The precise coordination of the central S phase events may thus rely on different strategies modulating histone protein production according to the metabolic needs of a cell. Finally, the emerged requirement of ABCE1 for histone biogenesis might converge with its highly conserved nature in various organisms, although a common mechanism for such function remains intriguing.
HEK293 and HeLa cells were cultured in Minimum Essential Medium (MEM, PAA) or High-Glucose (4.5 g/l) Dulbecco's modified Eagle's Medium (DMEM, PAA), respectively, containing 2 mM L-glutamine and 10% fetal bovine serum, at 37°C in a humidified atmosphere with 5% CO2. Cells were routinely passaged by trypsinization.
For siRNA transfection, 5 nM Silencer Select siRNAs from Ambion (ABCE1-specific: s12088 and s12089 referred here as ABCE1-1 and ABCE1-2, respectively; RNase L-specific: s12065 referred here as RNase L; and a negative control siRNA #2 referred here as Scrambled) and Lipofectamine® RNAiMAX Transfection Reagent (Life Technologies) were used as recommended by the manufacturer. In most ABCE1 depletion experiments, ABCE1-1 siRNA was applied. In the case of simultaneous transfection with two siRNAs, their final concentration was 5 nM each. In all experiments, cells were transfected by a reverse transfection procedure.
For cell growth estimation, cells were transfected on two parallel 96-w plates at confluence reaching about 70% on day 3 after transfection, at which time one of the plates was assayed. Cells from another plate were trypsinized, passaged at 1/3 dilution and assayed on day 6 after transfection. The number of vital cells was estimated by the MTT assay performed essentially as described by Wilson,105 with 0.5 mg/ml MTT (Sigma). The absorbance at 570 nm was measured with GENios Pro microplate spectrofluorometer (Tecan Group, Switzerland) equipped with the Magellan V5.03 system.
The efficiency of silencing was estimated by qPCR. Total RNA was extracted from cells using the TRIzol Reagent (Life Technologies) as recommended by the manufacturer, and 100 µg was reverse transcribed using the TaqMan reverse transcription reagents (Roche Molecular Systems). qPCR amplifications were performed in triplicates with 7900HT Fast instrument (Applied Biosystems, USA) using the TaqMan Gene Expression Assays for ABCE1 (Hs01003010_g1) and RPLO (large ribosomal protein) as endogenous control.
HEK293 cells were synchronized at the beginning of S phase using a double thymidine block. Cells were treated with siRNA for 3 days on a 9-cm plate and then seeded at equal densities on 6-cm plates. Next day, cells were incubated in medium with 2 mM thymidine for 16 h and subsequently washed and released for 8.5 h in fresh medium. Incubation with thymidine was repeated again for 15 h. On day 6 after transfection, cells were released from thymidine block by replacement of fresh medium. Cells were harvested at the time points indicated in figures, collected by centrifugation and fixed in ice-cold 70% ethanol. Cells were stained with 10 µg/ml propidium iodide in the presence of 100 µg/ml RNase A (Sigma) and analyzed using FACS Calibur flow cytometer (BD Biosciences, USA) and the CellQuest program.
For DNA synthesis studies, synchronized cells were pulse-labeled with 10 µM BrdU for 30 min during release. Cells were then harvested, fixed and treated with 2 M HCl for 30 min followed by neutralization with 0.1 M sodium tetraborate for 2 min. For BrdU detection, anti-BrdU mouse monoclonal antibody G3G4 (Developmental Studies Hybridoma Bank, University of Iowa, USA) at 1:50 dilution and Alexa 488-conjugated goat anti-mouse antibody (A11001, Invitrogen) at 1:1000 dilution were used. Thereafter cells were stained with propidium iodide and analyzed by flow cytometry as described above.
Histone mRNA levels in synchronized cells were estimated by qPCR. RNA was treated with DNase I (Thermo Scientific) following manufacturer's instructions. cDNA was synthesized from 150–200 ng of RNA using Maxima reverse transcriptase (Thermo Scientific) and random hexamer primer. qPCR was performed with LightCycler 480 II (Roche, Germany) instrument using 5x HOT FIREPol EvaGreen qPCR Mix Plus (Solis Biodyne). The reactions were performed in triplicates and target expression was normalized to the levels of β-actin mRNA. Primers for H2B and β-actin were previously described in Yu et al.80 and H4D primers were as follows: 5′- CTAAGCGCCACCGTAAAG-3′ and 5′-TAGATGAGGCCGGAGATG-3′.
For Western blotting, cells were lysed in RIPA buffer (50 mM Tris–HCl [pH 8.0], 150 mM NaCl, 1% NP-40, 0.5% sodium deoxycholate, 0.1% SDS) with protease inhibitors (Roche), incubated on ice for 10 min and centrifuged for 3 min at 14,000 × g. Protein concentrations were determined by the BCA assay (Thermo Scientific). Proteins were separated on 10 – 15% SDS-PAGE and electroblotted to nitrocellulose membrane (Hybond-C Extra, Amersham Biosciences). After blocking in 5% non-fat dried milk/0.1% Tween – PBS, the membranes were incubated with primary antibodies in 1% milk/0.1% Tween – PBS at following dilutions: ABCE1 1:800 (ab32270, Abcam and PAB18823, Abnova); actin 1:1000 (A 2066, Sigma); H2B 1:10000 (ab1790, Abcam); H4 1:5000 (ab10158, Abcam); SLBP 1:1000 (H00007884-M01, Abnova); eIF3A 1:2000 (NBP1-18891, Novus Biologicals); eIF3H 1:1000 (NBP1-31572, Novus Biologicals); RPA2 1:400 (ab16855, Abcam); RRM2 1:500 (H00006241-M01, Abnova); p27 1:100 (sc-528, Santa Cruz Biotechnology). HRP-conjugated secondary antibodies (32460 and 32430, Thermo Scientific) were used at 1:2500 - 1:5000 dilution. Signals were detected with SuperSignal West Femto Chemiluminescent Substrate kit (Thermo Scientific) and ImageQuant LAS 4000 imager (GE Healthcare, UK).
For protein de novo synthesis studies, cells were treated with siRNA for 3 days on 6-cm plates and then seeded on parallel plates. On day 6 after transfection, cells were pulse-labeled with 10 µCi/ml of [35S]-EasyTag™ EXPRESS Protein Labeling Mix (1000 Ci/mmol, Perkin Elmer) in methionine/cysteine-free DMEM (Sigma) for 30 min at 37°C. Cells were harvested, washed by centrifugation and lysed in 1% SDS with protease inhibitors. Proteins were separated on 10% SDS-PAGE and electroblotted to nitrocellulose membrane, which was autoradiographed by phosphorimaging (Molecular Imager FX, Bio-Rad, USA). The membrane was then incubated with antibodies as specified above. The cells from parallel plates were used for cell cycle analysis by flow cytometry in some experiments. Quantification of proteins was performed by densitometric analysis using Quantity One program (Bio-Rad).
Cells were grown on cover slips, fixed in 4% formaldehyde and permeabilized in 1% Triton X-100 in PBS followed by 3 washes with PBS after each step. The cells were then incubated with antibodies diluted in PBS as follows: rabbit polyclonal anti-ABCE1 (NB400-116, Novus Biologicals) 1:40 and Alexa 568-conjugated goat anti-rabbit IgG (A-11036, Molecular Probes) 1:750 followed by 3 washes with PBS after each incubation. Cell nuclei were stained with 1 µg/ml Hoechst 33342 dye (Molecular Probes) and the samples were analyzed by confocal microscopy (LSM Duo, Zeiss, Germany).
Statistical analysis was performed with Microsoft Excel and JMP10.0 (SAS) software. All data are expressed as mean ± standard deviation. Histone qPCR data and ABCE1 protein quantification data were analyzed as described in Willems et al.106 To calculate the two-sided p-values, Student's t test was used in the case of two groups and one-way ANOVA with Dunnett's post hoc tests was applied if there were more groups in one experiment.
No potential conflicts of interest were disclosed.
We thank Jaanus Remme for a critical reading of the manuscript and Heiti Paves for his help with confocal microscopy.
MT conceived the study, performed the experiments and wrote the manuscript; KK performed the experiments and contributed to the manuscript preparation; PP carried out the statistical analysis of the data; ET supervised the work and contributed to the manuscript preparation; CS conceived the study, supervised the work and contributed to the manuscript preparation. All authors approved the final manuscript.
The research was supported by the European Union through the European Social Fund (Mobilitas grant MJD121 to MT) and by institutional research funding IUT 193 of the Estonian Ministry of Education and Research.