P54nrb forms a heterodimer with PSP1 that localizes to paraspeckles in an RNA-dependent manner

Abstract

P54nrb is a protein implicated in multiple nuclear processes whose specific functions may correlate with its presence at different nuclear locations. Here we characterize paraspeckles, a subnuclear domain containing p54nrb and other RNA-binding proteins including PSP1, a protein with sequence similarity to p54nrb that acts as a marker for paraspeckles. We show that PSP1 interacts in vivo with a subset of the total cellular pool of p54nrb. We map the domain within PSP1 that is mediating this interaction and show it is required for the correct localization of PSP1 to paraspeckles. This interaction is necessary but not sufficient for paraspeckle targeting by PSP1, which also requires an RRM capable of RNA binding. Blocking the reinitiation of RNA Pol II transcription at the end of mitosis with DRB prevents paraspeckle formation, which recommences after removal of DRB, indicating that paraspeckle formation is dependent on RNA Polymerase II transcription. Thus paraspeckles are the sites where a subset of the total cellular pool of p54nrb is targeted in a RNA Polymerase II-dependent manner.

Figure 1.

Paraspeckles through the cell cycle. (A) Fluorescent micrographs of a HeLa^YFP-PSP1α cell acquired over the course of 16 h of live-cell imaging (see Materials and Methods). For movie see Supplementary Video1.mov. (B) Fluorescent micrographs of a representative live HeLa^YFP-PSP1α cell (YFP-PSP1, green signal) treated with bisbenzimide to label DNA (Blue signal), before imaging through mitosis. (C) DIC and fluorescence image of a HeLa^YFP-PSP1α cell at telophase, acquired as part of a live-cell time-course imaging experiment. Scale bar, 10 μm. Arrows, paraspeckles; arrowheads, perinucleolar caps.

Figure 2.

Transcription inhibition by DRB treatment prolongs the appearance of perinucleolar YFP-PSP1 in newly divided cells, with paraspeckles only forming once the drug is removed. A panel of fluorescent micrographs acquired by live-cell imaging HeLa^YFP-PSP1α cells every 3 min over the course of 7 h. At the start of the experiment, after cell division, DRB was added, and then removed, in this case, 1 h 22 min after cell division. Scale bar, 10 μm. Arrows, paraspeckles; arrowheads, perinucleolar caps. See Supplementary Video2.mov for movie.

Figure 3.

RNase A treatment abolishes YFP-PSP1α signal in paraspeckles, but not peri-nucleolar caps. Fluorescent micrographs of a representative HeLa^YFP-PSP1α cell (A and B) and a HeLa^YFP-PSP1α cell that was pretreated with ActD (C and D) following incubation with either buffer (A and C), or buffer containing RNase A (B and D). Blue, DAPI; green, YFP-PSP1. The absence of Pyronin Y staining was used to confirm the efficacy of the RNase A treatment (unpublished data). Scale bar, 10 μm. Arrows, paraspeckles; arrowheads, perinucleolar caps.

Figure 4.

Regions of PSP1 required for localization to paraspeckles and perinucleolar caps. The top panel shows a summary of the localization pattern of YFP-fusions of PSP1 mutants expressed in HeLa cells. Below, representative fluorescence micrographs of a subset of the YFP-PSP1 mutants expressed in HeLa (green signal) and immunostained with a paraspeckle marker (red signal) that was either PSP1 (PSP1 Ab2 for mutants A, B, D, E, F, G, and K), or PSF (monoclonal B92, YFP-PSP1 and mutants C and H). The choice of paraspeckle marker was dictated by the location of the peptide epitope the PSP1 Ab2 was raised against (at the extreme C terminus of PSP1); therefore any PSP1 mutant that contained this region could not be compared with endogenous PSP1 due to antibody cross-reactivity with the overexpressed fusion proteins. Representative fluorescent micrographs of mutant localization patterns following ActD treatment are shown in Supplementary Figure 2. Scale bars, 10 μm. “Paraspeckle” localization is defined as the formation of multiple nucleoplasmic foci colocalizing with a paraspeckle marker protein with a fluorescence intensity quantitatively more than twice the background nucleoplasmic signal. “cap” localization is defined as the presence of perinucleolar caps after treatment of transfected cells with ActD (1 μg/ml) for 4 h. Amino acid substitution mutations are indicated by an × and are F119A, F121A (for RRM1) and K198A, F200A (for RRM2).

Figure 5.

PSP1 associates with p54nrb in vivo. (A) Endogenous PSP1 was immunoprecipitated from HeLa nuclear extract using two different PSP1 antibodies (preimmune, lane 2; Ab1, lane 3; and Ab2, lane 4). The IP proteins were run on a 4-12% SDS-PAGE and stained with colloidal blue. (B) To elute specific PSP1 complexes, IP-bound beads were incubated with buffer (lanes 2 and 4), or the cognate peptide that each antibody was raised against (Ab1 peptide, lane 3 and Ab2 peptide, lane 5). The supernatant was subjected to SDS-PAGE, and the gel silver-stained. (C) IP PSP1 complex (using Ab2) was repeatedly washed with 0.1-1 M NaCl/0.1-2% NP40 (see Materials and Methods). The bound proteins were subjected to SDS-PAGE and stained with colloidal blue. In all panels, molecular weight markers are shown in lane 1 (kDa). (D) IP with anti-GFP antibodies using HeLa whole cell lysates transiently expressing either YFP-p54nrb (lane 2), YFP-PSP1 (lane 3), or YFP (lane 4). The resultant bound proteins were separated by SDS-PAGE and immunoblotted with anti-PSP1 antibody. “input” (lane 1) is HeLa whole cell lysate.

Figure 6.

PSP1 requires its coiled coil domain to pull down p54nrb. Whole cell lysates of cells transiently expressing either YFP (lane 1), YFP-PSP1α (lane 2), or mutations in YFP-PSP1α (lanes 3-13) were used for IP with anti-GFP antibody-cross-linked beads. The resultant bound proteins were then subjected to SDS-PAGE, transferred to nitrocellulose and immunoblotted with p54nrb mAb (top panel), or GFP mAb (bottom panel), using ECL plus to visualize the signal.

Figure 7.

PSF localizes to paraspeckles. The panels show fluorescence micrographs of HeLa ^YFP-PSP1α cells (YFP fluorescence is green signal; A and C), or HeLa cells transiently expressing GFP-PSF (GFP fluorescence is green signal; B and D). Top panels (A and B) are untreated cells and bottom panels are cells incubated with ActD (C and D; see Materials and Methods). In each case, cells were immunostained with anti-PSF (red signal; A and C), or anti-PSP1 (red signal; B and D). Arrows, paraspeckles; arrowheads, perinucleolar caps. Blue signal is DAPI staining in all panels. Scale bar, 10 μm.

Figure 8.

PSP1 and p54nrb associate when RNA Pol II transcription is inhibited and form a heterodimer in vitro. (A) Western blot of HeLa nuclear extract before (lanes 1 and 3) and after IP (lanes 2 and 4) of endogenous PSP1 and its partner proteins (lanes 1-4) and the immunoprecipitated proteins bound to the beads (lanes 5-8). Extracts were prepared from control HeLa cells (lanes 1, 2, 5, and 6) and HeLa cells treated with ActD (lanes 3, 4, 7, and 8). To monitor that ActD treatment was effective, the relocalization of PSP1 to perinucleolar caps was checked before cell lysis. Either preimmune serum (lanes 5 and 7) or PSP1 Ab2 were used for IP (lanes 6 and 8). The protein extracts and washed beads were subjected to SDS-PAGE, transferred to nitrocellulose and Western blotted with either anti-PSP1 (top panel), or anti-p54nrb (bottom panel). The signal was detected with ECL plus. (B) Lysates of E. coli transformed with pETDuet-His₆-PSP1(61-328)/p54nrb(75-313) and induced for protein expression (see Materials and Methods) were loaded on Ni-agarose, the flow-through was collected (lane 1), the column was washed (lane 2), and fractions were eluted with increasing imidazole (lanes 3-8). Samples were subjected to SDS-PAGE and stained with Colloidal blue. Molecular weight markers, lane 9 (kDa).

Figure 9.

Differing dynamics of fluorescent-PSP1 in paraspeckles versus perinucleolar caps. The graph shows the combined data (mean and 95% confidence interval) of the loss of photoactivated paGFP-PSP1 protein from paraspeckles and perinucleolar cap structures in 30 different HeLa cells transiently expressing PA-GFP-PSP1 (see Materials and Methods). For each time point, the background (preactivation) fluorescence was subtracted, and the resultant value was expressed as a ratio of the first activated fluorescence value (i.e., value at 0 s). The data points are connected by a curve of best fit (2-phase exponential decay).