Phosphorylation regulates coilin activity and RNA association

Coilin hyperphosphorylation correlates with a decrease in its association with U2 snRNA and hTR

We have found (Broome and Hebert, 2013) that coilin associates with specific RNA and DNA. In particular, we have observed that coilin immunoprecipitation complexes contain a 3.6-fold enrichment of U2 snRNA and a 17.5-fold enrichment of hTR relative to control IgG reactions in untreated HeLa cells (Broome and Hebert, 2013). These untreated data are also shown in Fig. 1 for reference. To determine if the phosphorylation of coilin alters the amount of RNA found in the coilin immunoprecipitation complex, nocodazole was used to enrich for cells in mitosis. Coilin is a phosphoprotein, and the amount of this phosphorylation increases during mitosis (Carmo-Fonseca et al., 1993). Evidence from MS/MS analysis indicates that at least four residues, in addition to those that are phosphorylated during interphase, are phosphorylated specifically during mitosis (Beausoleil et al., 2004; Dephoure et al., 2008; Toyota et al., 2010). Relative to untreated cells, therefore, coilin is hyperphosphorylated in nocodazole treated cells. For these experiments, lysate was immunoprecipitated with control (purified rabbit IgG) or coilin antibodies, followed by extensive washing of the beads and RNA isolation. The isolated RNA was then subjected to reverse transcriptase real-time PCR using primers specific to RNAs known to accumulate in the CB (U1 snRNA, U2 snRNA and hTR). Given that certain conditions, such as DNA damage (Gilder et al., 2011), result in the relocalization of coilin to the nucleolus, we also examined the association of coilin with the RNA pol I derived pre-processed (47/45S) and processed 5.8S rRNAs. Associated GAPDH message level was used as a control. Immunoprecipitation complexes containing hyperphosphorylated coilin contain significantly less U2 snRNA and hTR than that recovered from untreated cells (Fig. 1, nocodazole), demonstrating that increased coilin phosphorylation correlates with a reduction in the association of coilin with these RNAs. In contrast, the amount of pre-rRNA (47/45S) recovered in the coilin complex does not change after nocodazole treatment.

Transient expression of coilin wild type or S489 phosphomutants differentially impact the level of specific RNAs

Transient expression of coilin wild type (WT) and various phosphomutants demonstrate that mutation of the S489 phosphorylation site to aspartic acid (S489D), which mimics constitutive phosphorylation, results in the reduction of cell proliferation compared to cells with ectopically expressed WT coilin (Carrero et al., 2011). The phosphorylation of coilin S489 is enriched during mitosis (Beausoleil et al., 2004), indicating that this may be a mechanism for regulation of coilin activity. Since we have found that coilin directly binds RNA and DNA (Broome and Hebert, 2012), and can associate with U2 gene loci, as assessed by chromatin immunoprecipitation (ChIP) (Broome and Hebert, 2013), we decided to test if the transient expression of GFP-coilin (S489D) alters the message levels of specific RNAs. Towards this end, cells were transfected with GFP-tagged WT, phosphomimic (S489D) or phosphonull (S489A) coilin constructs, followed by RNA isolation and detection of messages by reverse transcriptase real-time PCR (qRT-PCR). The level of RNAs known to localize to the CB, such as U1 and U2 snRNA and hTR, as well as the pre-processed forms of these transcripts, were examined. As mentioned above, RNA pol I derived transcripts (the pre-processed 47/45S and the processed 5.8S rRNAs) were also examined since coilin nucleolar relocalization is observed in certain stress conditions. The levels of these various RNAs were normalized to the amount of GAPDH in cells expressing GFP. To our surprise, ectopic expression of GFP-coilin WT significantly decreased the levels of all RNAs tested except for U2 snRNA, relative to GAPDH and normalized to cells expressing GFP alone (Fig. 2). We have shown previously that GFP-coilin WT overexpression results in the decrease of both hTR and pre-hTR levels (Broome and Hebert, 2013), and this result was also found in this study (Fig. 2). We then tested if the transient expression of S489 phosphomutants would likewise decrease specific RNA messages as found for WT coilin expression. For the most part, message levels in cells expressing GFP-tagged S489D or S489A were reduced in a similar trend as observed for cells expressing GFP-coilin (Fig. 2). However, there are a few notable differences. The most striking of which is the decrease observed in the level of mature U2 snRNA in cells expressing GFP-coilin (S489A) compared to the relative increase of this snRNA in GFP-coilin or GFP-coilin (S489D) expressing cells. Regarding S489D expression, the amount of pre-U1 is not reduced by this protein to the same extent as that found upon the expression of WT or S489A coilin. When comparing RNA levels from cells expressing GFP-coilin to that found in cells expressing S489D or S489A, WT coilin is most divergent from the phosphomutant proteins in the level of U1 snRNA and pre-hTR. In summary, these results demonstrate that coilin overexpression reduces the level of several RNAs known to accumulate in the CB or associate with coilin. The extent of this change varies in cells expressing S489D or S489A phosphomutant proteins. Taken together, these findings suggest that coilin overexpression decreases the transcription of these genes and/or increases the degradation or processing of these specific RNAs, and that mimicking phosphorylation of serine 489 alters this profile for certain transcripts.

Coilin knockdown results in transcription and/or processing consequences in both transformed and primary cell lines

In addition to exploring how the presence of ectopic coilin affects RNA expression and processing, we examined what effect the transient knockdown of endogenous protein would have on these transcripts. We chose to examine total U2 snRNA, pre-U2 snRNA, 47/45S rRNA and 5.8S rRNA given their association with coilin and/or altered levels upon coilin overexpression. We utilized both HeLa and WI-38 cell lines for these experiments due to the fact that they differ not only in terms of transformed/non-transformed status but also in the number and prominence of CBs. While HeLa cells contain 2–6 CBs per cell (Leeflang et al., 1992), only 2–3% of WI-38 cells have a single CB (Spector et al., 1992). Cells were transfected with either a control non-targeting siRNA or one targeting coilin mRNA for 48 hours, and then the RNA was analyzed by qRT-PCR using specific primers. We have previously shown that overexpression of coilin results in a significant decrease in relative 47/45S pre-rRNA levels in HeLa cells (Gilder et al., 2011), and this result was also observed in this study (Fig. 2). In Fig. 3A, we show that following a 48-hour knockdown of coilin in HeLa cells, there are no significant changes in the levels of either 5.8S rRNA or 47/45S pre-processed rRNA. We have previously published that coilin knockdown in HeLa cells results in a significant decrease in total U2 snRNA and an increase in the pre-processed U2 transcript (Broome and Hebert, 2012). Also, we have found that coilin knockdown in HeLa cells results in a significant decrease in pre-hTR, and that coilin overexpression results in a significant decrease in both total and pre-processed hTR (Broome and Hebert, 2013). In Fig. 3B, we show the effect of coilin knockdown on U2 snRNA, hTR and rRNA levels in WI-38 cells. While coilin knockdown resulted in a nearly 1.6 fold increase in total U2 levels, there was essentially no change in the levels of the pre-processed U2 transcript. This is an interesting finding due to the fact that an opposite trend is seen in HeLa cells following coilin knockdown (Broome and Hebert, 2012), in which there is an increase of pre-processed and a small decrease of total U2 snRNA. In agreement with what we find in HeLa cells (Broome and Hebert, 2013), we find a slight decrease in total hTR with a 57% decrease in pre-hTR upon coilin knockdown in WI-38 cells. In addition, we find a 1.9 fold increase in the 47/45S pre-rRNA with no change in 5.8S rRNA levels. Interestingly, the relative amount of coilin message is 2.5 fold greater in WI-38 versus HeLa (Fig. 3C). Collectively, these data suggest a role for coilin in the transcription or processing of specific RNAs in cell types with or without CBs.

Bacterially purified coilin (S489D) co-isolates with more nucleic acid than WT coilin

In addition to nucleic acid binding, we have found that coilin has RNase activity (Broome and Hebert, 2012). To further explore the role of S489 phosphorylation on coilin function, with particular emphasis on how this modification may impact the RNase activity of coilin, we purified GST-coilin (S489D) from bacteria using a stringent purification protocol involving removal of the GST-tag and excision of the full-length protein from a SDS-PAGE gel followed by electro-elution (Broome and Hebert, 2012). The RNase activity of S489D when compared to WT coilin was initially found to be equivalent when using total RNA as the substrate (Fig. 4A). Interestingly, the amount of co-purified nucleic acid, which we have previously shown to be mostly RNA (Broome and Hebert, 2012), is greater in S489D compared to WT (Fig. 4B). In order to more definitively characterize the RNase activity of the S489D phosphomutant compared to WT coilin, we utilized a modified coilin purification protocol (Broome and Hebert, 2013) in which nucleic acid free protein is isolated. In brief, protein samples are treated with DNase I and RNase A/T1 before SDS-PAGE and band excision. After electro-elution, the samples are treated with α-cyclodextrin (to strip SDS from the protein) followed by extensive dialysis in PBS containing 250 mM NaCl. The end result of this modified protocol is soluble purified coilin lacking any nucleic acid. Both WT and S489D proteins purified in this manner were subjected to RNA degradation assays using total RNA (Fig. 4C). Both proteins appear to degrade RNA approximately equally, although quantification of the 28S rRNA band demonstrates that there is a very slight reduction of RNase activity in the S489D phosphomutant compared to WT (Fig. 4D).

Differential processing of U2 snRNA by coilin WT, S489D and the deletion mutant d121–291

Our previous work demonstrates that in vitro transcribed U2 is selectively degraded by purified coilin, suggesting that coilin may take part in the 3′ end processing of this snRNA (Broome and Hebert, 2012). Specifically, coilin preferentially degraded the CU region of the extended U2 snRNA transcript. Since the coilin protein used in the previous study also contained co-purifying nucleic acid, we wanted to test if purified coilin lacking nucleic acid also shows a similar pattern of specificity. We also tested the activity of the S489D phosphomutant (likewise purified in the absence of nucleic acid) on the processing of an U2 snRNA substrate. For these studies, an 820 nt in vitro transcribed U2 snRNA, containing an extended 3′ that encompasses the CU (CT) box, was used as a substrate for reactions containing buffer or purified nucleic-acid free WT or S489D coilin proteins. After incubation, the reaction mix was DNase treated followed by qRT-PCR and agarose gel electrophoresis. The primers used for qRT-PCR amplify the 5′ end containing the mature region, the middle CU region, and the extreme 3′ end (Fig. 5A). When compared to buffer alone, a 2:1 (nM) ratio of protein to U2 snRNA revealed little degradation of the 5′ end as assessed via reverse transcriptase qRT-PCR (Fig. 5B). In sharp contrast, the amount of the U2 snRNA CU region is reduced 94% in reactions containing WT and 70% for reactions with S489D, relative to control. The U2 snRNA 3′ end is reduced by both WT and S489D proteins to 50% of that found in control reactions. A portion of the same RNase reactions used for the qRT-PCR set-up were also run on an agarose gel, and it is clear that the WT protein shows more degradation of the U2 snRNA substrate while the RNA degradation activity of the S489D protein is relatively blunted (Fig. 5C).

We have previously identified a deletion mutant of coilin, lacking amino acids 121–291 (d121–291), that binds less RNA and has less total RNA degradation activity compared to WT coilin (Broome and Hebert, 2013). Since the S489D mutant has reduced degradation of the U2 snRNA CU region compared to WT, we next wanted to test if the coilin (d121–291) mutant would also show similar impaired U2 snRNA degradation. For these studies, GST, GST-coilin and GST-coilin (d121–291) were purified using the modified protocol (Broome and Hebert, 2013) such that nucleic acid free, dialyzed protein was isolated. Note that for these studies the GST tag was left fused to the WT and d121–291 coilin proteins. The same 820 nt U2 snRNA substrate used above was incubated with GST, GST-coilin or GST-coilin d121–291, followed by DNase treatment, qRT-PCR and agarose gel electrophoresis. Relative to the reactions containing GST, WT coilin only modestly reduces the U2 5′ end, while the d121–291 protein has no effect (Fig. 5D). However, the CU region was reduced 80% in WT reactions, and 37% in d121–291 reactions. The 3′ end was reduced by both proteins, but more so for WT compared to d121–291. The same reactions used as the template for qRT-PCR were also run on an agarose gel for visualization (Fig. 5E), and it is clear that relative to GST, WT coilin is more effective at degrading the U2 snRNA substrate. In contrast, and in agreement with that observed with the S489D mutation, the RNA degradation activity of the d121–291 protein for the U2 snRNA substrate is reduced compared to WT.

Mutation of S489 and deletion of amino acids 121–291 impair coilin degradation of hTR

In the context of total RNA, incubation with purified WT coilin specifically reduces the amount of the 3′ end of hTR (Broome and Hebert, 2013), which is removed during pre-hTR processing in order to become the mature hTR that is incorporated into telomerase. In contrast, incubation of total RNA with purified coilin d121–291 does not result in reduced hTR 3′ end levels (Broome and Hebert, 2013). To examine if the coilin S489D and d121–291 proteins have reduced degradation activity using hTR as a substrate, as observed with the U2 snRNA substrate as described above, we conducted experiments with hTR that contains an extended 3′ end. Primers were used that amplify the mature region and the 3′ end spanning the processing site (Fig. 6A). For the first set of experiments, substrate was incubated with buffer or nucleic acid free WT or S489D coilin proteins, followed by DNase treatment. The reactions were then used as a template for reverse transcriptase qRT-PCR or were subject to agarose gel electrophoresis. As seen in Fig. 6B, the amount of both the mature hTR and 3′ end was reduced 75% in reactions containing WT coilin compared to control. As found previously for the CU region of the U2 snRNA substrate, this reduction was not as drastic in reactions containing coilin S489D (Fig. 6B). Visualization of the reactions by agarose gel confirms this observation (Fig. 6C).

To test if the d121–291 mutant will again share a reduced degradation phenotype with S489D using the hTR substrate, reactions were conducted using purified GST, GST-coilin or GST-coilin d121–291. While the amount of both the mature hTR and 3′ end were reduced upon incubation with WT coilin compared to GST, the level of reduction was not as severe in reactions containing d121–291 (Fig. 6D). Agarose gel electrophoresis of the same samples that were used as the template for the qRT-PCR reactions corroborates this observation (Fig. 6E). In summary, therefore, differential processing is observed for hTR upon coilin incubation in the context of total RNA (Broome and Hebert, 2013). However, when using only hTR RNA in a reaction, coilin appears to degrade both the 5′ and 3′ ends without preference. Interestingly, as observed for the U2 snRNA substrate, the S489D and d121–291 coilin mutants both display reduced RNase activity for the hTR substrate compared to WT, with the d121–291 protein being the least active.

Cell culture and DNA constructs

Human cervical carcinoma (HeLa) and normal human fetal lung fibroblast (WI38) cells were purchased from the American Type Culture Collection (Manassas, VA, USA). Cell lines were cultured as described previously (Sun et al., 2005), and GFP expression vectors have been described (Hebert and Matera, 2000). GST-coilin wild types (GST-WT and WT) were cloned into pGEX vectors as described previously (Broome and Hebert, 2012). GST-Δ121–291 was generated via the Quick Change Mutagenesis kit from Stratagene (Santa Clara, CA, USA) using pGEX-2T-coilin as template with the following primers: 5′-ATCGTCGCAGGATCCGACAAACTGGCTATAAAACTTGGC-3′ and 5′-GGATCCTGCGACGATTTCACCCTCCTCTAACTGAAATGC-3′. GST-coilin S489D was generated using the Quick Change Mutagenesis kit from GST-coilin wild type. GFP-coilin S489D and GFP-coilin S489A phosphomutants were generated by Quick Change Mutagenesis (Carrero et al., 2011) from a GFP-coilin WT precursor (Hebert and Matera, 2000). pBluescript KS U2 used for generation of the 820 nt U2 RNA transcript has been described previously (Broome and Hebert, 2012). The pBluescript KS hTR expression vector was generated by first amplifying HeLa genomic DNA with the following primers that contain engineered sites for SacI and EcoRI (underlined) designed by using the human TERC NG_016363.1 sequence: 5′-GCCGAGCTCCGGGTTGCGGAGGGTGGGCCTGGGAG-3′ and 5′-GCCGAATTCCCATTGCCGGCGAGGGGTGACGGATG-3′. Following amplification, the PCR product was cloned into a PCR4-TOPO vector using the TOPO TA Cloning kit from Invitrogen (Carlsbad, CA, USA). The SacI/EcoRI digested hTR insert was then cloned into SacI/EcoRI digested pBluescript KS vector. The sequence was verified against NG_016363.1 and found to contain one C to T mutation at nucleotide 5087 of NG_016363.1.

Transfections, RNAi and drug treatment

Cells were transfected with GFP expression vectors using Fugene HD Transfection Reagent (Promega, Madison, WI, USA) following the manufacturer's suggested protocol. Coilin siRNA (Toyota et al., 2010) and the non-targeting control siRNA (Carrero et al., 2011) have been described previously. The siRNA transfections were carried out using Lipofectamine 2000 (Invitrogen, Carlsbad, CA, USA) according to the manufacturer's protocol. Nocodazole treatment was for 16 hours at 0.4 µg/ml.

RNA IP

Approximately 20 million cells were lysed in 1 ml RIPA buffer (50 mM Tris-HCl pH 7.6, 150 mM NaCl, 1% NP-40, 0.25% sodium deoxycholate, 0.1% SDS, 1 mM EDTA) containing protease inhibitor cocktail (Roche, Indianapolis, IN, USA) for RNA IP experiments. Cell lysates were sonicated four times for 5 seconds each at 4 W using the Fisher Scientific Sonic Dismembrator Model 100, then centrifuged at 16000 g for 5 minutes at 4°C. Sonicate from ∼20 million cells was used for each RNA IP reaction. Antibody and 45 µl 50% Protein G Sepharose beads (GE Healthcare, Pittsburgh, PA, USA) were added to the sonicate and incubated at 4°C for ∼18 hours on a multi-tube rotator. Normal rabbit IgG (sc-2027) and α-coilin H300 rabbit polyclonal (sc-32860) antibodies were purchased from Santa Cruz Biotechnology (Santa Cruz, CA, USA); 5 µg of each antibody was used per IP. Following the IP incubation, antibody/bead complexes were pelleted at 16000 g for 1 minute, the supernatant removed, and complexes washed in 1 ml RIPA buffer 3 times. RNA was then isolated from the antibody/bead complex using a modified Ambion RNAqueous kit protocol (Life Technologies, Carlsbad, CA, USA). Briefly, 350 µl lysis buffer and 350 µl 64% ethanol was added to the aspirated antibody/bead complex and the subsequent steps performed per the manufacturer's suggested protocol. RNA was eluted in 100 µl. Following elution, samples were incubated at 37°C for 30 minutes with TURBO DNase (Life Technologies, Carlsbad, CA, USA), then equal volumes of isolated RNA were used for the qRT-PCR analysis.

Bacterially expressed proteins

GST-tagged coilin proteins containing co-isolated nucleic acid were purified to SDS-PAGE single band homogeneity as described previously (Broome and Hebert, 2012). For nucleic acid-free proteins, the protocol was modified as follows: protein was incubated at 37°C for 30 minutes with DNase I and RNase A/T1 cocktail before separation by SDS-PAGE. Following electro-elution from the gel slices, the protein was put over a Pierce detergent removal column (Thermo Scientific (Waltham, MA, USA) for SDS-removal then concentrated using a centrifugal filter unit with 50K molecular weight cutoff (Millipore, Billerica, MA, USA). The protein was then incubated on ice for 30 minutes at a final concentration of 21 mM α-cyclodextrin (Otzen and Oliveberg, 2001) and 250 mM NaCl in order to remove residual protein-bound SDS, followed by centrifugation for 5 minutes at 16000 g. The resulting supernatant was then dialyzed against 1× PBS high salt buffer containing 10 mM Na₂HPO₄, 1.8 mM NaH₂PO₄ and 250 mM NaCl using a 10 K MWCO Pierce Slide-A-Lyzer Dialysis Cassette from Thermo Scientific following the manufacturer's suggested protocol. Concentration of dialyzed proteins was estimated using Pierce BCA Protein Assay Kit (Thermo Scientific, Waltham, MA, USA) following the manufacturer's protocol.

RNase assays

RNase assays for agarose gel end point analysis were conducted with 500 ng HeLa RNA isolated using the Macherey–Nagel NucleoSpin RNA II kit (Clontech, Mountain View, CA, USA) and up to 500 ng purified protein. Reactions were incubated at 37°C for 30 minutes, loaded into a 1% agarose gel and visualized by ethidium bromide. Densitometry of gel bands was measured using Quantity One software. U2 and hTR transcripts were generated from their respective pBluescript vectors using the MaxiScriptT7 kit (Ambion Life Technologies, Carlsbad, CA, USA) per the manufacturer's suggested protocol. RNase assays with U2 and hTR transcription reactions were conducted with 100 nM RNA and either 200 nM (300 ng) coilin WT or S489D or 300 ng GST-WT or GST-d121–291 protein at 37°C for 30 minutes. Following incubation, reactions were treated with TURBO DNase (Ambion Life Technologies, Carlsbad, CA, USA) for 20 minutes at 37°C. Equal volumes of the reactions were used for qRT-PCR.

qRT-PCR

Brilliant II SYBR Green qRT-PCR Master Mix Kit from Agilent (Santa Clara, CA, USA) was used for analysis of RNA from RNA IP experiments, transfected cells and U2 and hTR RNase assays and the manufacturer's suggested thermal cycling parameters, primer concentrations and RNA amounts were used. Primers for GAPDH, 47/45S pre-rRNA, U1, pre-U1, U2 and pre-U2 have been described (Broome and Hebert, 2012; Gilder et al., 2011; Hearst et al., 2009). Primer sequences are: 5.8S rRNA: 5′-CGGCTCGTGCGTCGAT-3′ forward and 5′-CCGCAAGTGCGTTCGAA-3′ reverse; telomerase RNA (hTR): 5′-AAATGTCAGCTGCTGGCCCGTTCG-3′ forward and 5′-ACCCGCGGCTGACAGAGCCCAAC-3′ reverse; pre-processed telomerase RNA (pre-hTR): 5′-AGGTTCAGGCCTTTCAGGCCGCAG-3′ forward and 5′-GACGGATGCGCACGATCGGCGTTC-3′ reverse. The Stratagene Mx3000P Real-Time PCR System was used with Agilent MxPro software for real-time analysis using Windows Excel for post hoc statistical analysis. The Student's t-test for independent variables with unequal variance was used for calculation of P-values.