Coilin displays differential affinity for specific RNAs in vivo and is linked to telomerase RNA biogenesis

Abstract

Coilin is widely known as the protein marker of the Cajal body, a subnuclear domain important to the biogenesis of small nuclear ribonucleoproteins and telomerase, complexes that are crucial to pre-messenger RNA splicing and telomere maintenance, respectively. Extensive studies have characterized the interaction between coilin and the various other protein components of CBs and related subnuclear domains; however, only a few have examined interactions between coilin and nucleic acid. We have recently published that coilin is tightly associated with nucleic acid, displays RNase activity in vitro, and is redistributed to the ribosomal RNA (rRNA)-rich nucleoli in cells treated with the DNA-damaging agents cisplatin and etoposide. Here, we report a specific in vivo association between coilin and rRNA, U small nuclear RNA (snRNA), and human telomerase RNA, which is altered upon treatment with DNA-damaging agents. Using chromatin immunoprecipitation, we provide evidence of coilin interaction with specific regions of U snRNA gene loci. We have also utilized bacterially expressed coilin fragments in order to map the region(s) important for RNA binding and RNase activity in vitro. Additionally, we provide evidence of coilin involvement in the processing of human telomerase RNA both in vitro and in vivo.

Figure 1

(A) Histogram of RNA IP results with IgG and α-coilin from untreated, etoposide or cisplatin treated HeLa cells; fold enrichment of α-coilin IP over IgG control is shown on log₁₀ scale; error bars represent 1 s.d. for 3 experimental repeats with 3 technical repeats each; * indicates p<0.005 relative to GAPDH within a treatment; # indicates p<0.002 relative to untreated cells within a primer set; ✠ indicates p<0.04 between etoposide and cisplatin within a primer set. (B) Histogram of RNA IP results with IgG, α-fibrillarin and α-SMN from HeLa cells; fold enrichment over IgG control is shown on log₁₀ scale; error bars represent 1 s.d. for 1–3 experimental repeats per primer and 3 technical repeats per experiment; * indicates p<0.05 and ** p<0.001 relative to GAPDH within an IP; # indicates p<0.02 between IPs for the same primer set. (C) Histogram of RNA IP results with IgG and α-coilin from WI-38 cells; fold enrichment of α-coilin over IgG is shown; error bars represent 1 s.d. for 2 experimental repeats with 3 technical repeats each; * indicates p<0.01 relative to GAPDH.

Figure 2

(A) Diagram showing primer binding locations within the target regions of the U1 and U2 snRNA gene loci. (B) Histogram of ChIP results using antibodies to coilin and RNA polymerase II in HeLa cells, fold enrichment of α-coilin and α-RNA polymerase II over IgG control is shown; * indicates p<0.005 relative to IgG; # above data set for a primer pair indicates p<0.01 between coilin and RNA polymerase II IPs. Error bars represent 1 s.d. for 3 experimental repeats with 3 technical repeats each.

Figure 3

(A) Diagrams of seven GST-tagged coilin constructs showing precipitation behavior of equal total protein, 160 μg, following incubation with or without DNase and RNase, with “-” being no visible precipitation and “++++” the most relative precipitate; SA = self-association, FL = full length. (B) Top panel, Coomassie stained SDS-PAGE gel of equal total partially purified GST-tagged proteins untreated; Bottom panel, equal volume of soluble protein following incubation with DNase and RNase; * indicates full length GST-tagged protein of interest for each construct; % Soluble value indicates the percentage of soluble protein remaining after nuclease treatment, relative to untreated soluble protein, as calculated by relative desitometric analysis of protein bands in top and bottom panels. (C) Equal amounts of total GST-tagged protein loaded into an agarose gel and stained with ethidium bromide to visualize relative amounts of co-purified nucleic acid. (D) Representative image of visible precipitate which forms upon nuclease treatment, seen with RNase treatment or the combination of DNase and RNase.

Figure 4

(A) Coomassie stained SDS-PAGE gels containing fully purified nucleic acid-free GST-tagged proteins; ladder to the right of each separate panel denotes marker locations of 175, 80, 58, 46, 30 and 25 kDa for each individual gel. (B) Agarose gels containing RNase assay results visualized by ethidium bromide; 500 ng RNA per lane; numbers beneath lanes indicate amount of protein in μg; the 28S rRNA band is indicated, with both the 32S and 18S also visible. (C) Line graph showing densitometric analysis of the 28S rRNA bands from images shown in B; percent band density of the control (no protein) band is plotted for each protein and amount.

Figure 5

(A) RNase activity of nucleic acid free purified coilin with 500 ng RNA; numbers beneath lanes indicate amount of protein in μg. (B) Histogram showing the fold change relative to control reactions of hTR and pre-hTR levels following incubation of HeLa RNA with nucleic acid free purified coilin; * indicates p<0.05 relative to the control incubation.(C) Histogram showing the fold change relative to control reactions of hTR and pre-hTR levels following incubation of HeLa RNA with nucleic acid free purified GST-tagged coilin proteins;* indicates p<2.0E-9 relative to GST-coilin d121-291 and GST-coilin C214. Error bars represent 1 s.d. for 3 experimental repeats with 3 technical repeats each.

Figure 6

(A) Histogram showing the fold change relative to GFP transfection of hTR and pre-hTR levels following a 24h transfection with GFP-coilin wt; # indicates p<2E-6 and * indicates p<2E-8 relative to total hTR. (B)Histogram showing the fold change relative to control knockdown of hTR and pre-hTR levels following a 48h coilin knockdown in HeLa cells resulting in 89% reduction of coilin mRNA; # indicates p<6E-5 relative to control KD and * indicates p<0.001 relative to total hTR. Error bars represent 1 s.d. for 3 experimental repeats with 3 technical repeats each.