![[Feedback]](/icons/banner/feedbackACT.gif){kind=icon}
![[For Subscribers]](/icons/banner/subscriberACT.gif){kind=icon}
![[Archive]](/icons/banner/archiveACT.gif){kind=icon}
![[Search]](/icons/banner/searchACT.gif){kind=icon}
![[Contents]](/icons/banner/tocACT.gif){kind=icon}
|
|
|
|
|
||
Vol. 16, Issue 6, 2862-2871, June 2005
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
* Nuclei and Cell Cycle Laboratory, Institut Jacques Monod, Centre National de la Recherche Scientifique, University Paris VI and Paris VII, 75251 Paris, France;
Macromolecular Complexes in Live Cells, Institut Jacques Monod, Centre National de la Recherche Scientifique, University Paris VI and Paris VII, 75251 Paris, France; and
Microtubules and Morphogenesis Laboratory, Ecole Supérieure de Biotechnologie de Strasbourg, Centre National de la Recherche Scientifique Unité Mixte Recherche 7100, 67400 Illkirch-Graffenstaden, France
Submitted January 19, 2005;
Revised March 18, 2005;
Accepted March 24, 2005
Monitoring Editor: Joseph Gall
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONACKNOWLEDGMENTSREFERENCES |
|---|
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONACKNOWLEDGMENTSREFERENCES |
|---|
; Lamond and Earnshaw, 1998; Matera, 1999; Gall, 2000; Misteli, 2000; Spector, 2001; Hernandez-Verdun et al., 2002; Huang, 2002; Lamond and Spector, 2003). This implies that the recruitment of dedicated machineries and formation of discrete nuclear domains are crucial events at the beginning of interphase. In addition, recent data indicate that the entry as well as correct relocation of the nuclear machineries into newly forming nuclei is a step-dependent and ordered process (Savino et al., 2001; Prasanth et al., 2003; Bubulya et al., 2004; Leung et al., 2004). However, recruitment of the nucleolar processing proteins on transcription sites is still a poorly understood process. To better understand the rebuilding of nucleolar functions after mitosis, we chose to investigate the assembly of the nucleolar processing machinery in real time in living cells.
The nucleolus is a functional compartment of the nucleus generated by ribosome biogenesis (Hadjiolov, 1985; Mélèse and Xue, 1995; Leung et al., 2003). It is also a plurifunctional domain involved in the three-dimensional (3D) organization of chromatin in the nucleus (Chubb et al., 2002), in the assembly of several ribonucleoprotein complexes (Pederson, 1998; Olson et al., 2000), and in the formation of nuclear speckles (Bubulya et al., 2004). The nucleolus therefore seems to be a key player of the nuclear functional architecture. Its functions depend on the activation and recruitment of the nucleolar machineries involved in transcription of the ribosomal genes (rDNA) and processing of ribosomal RNAs (rRNA). These machineries are inherited through mitosis from the previous interphase (Hernandez-Verdun et al., 2002). During mitosis, the nucleolar machineries that participate in rRNA processing within the nucleolus are distributed around all the chromosomes and therefore at some distance from the rDNA sequences. When rDNA transcription is activated during telophase, the processing machineries are targeted to the sites of rRNA synthesis. Along the translocation pathway between chromosome periphery and sites of transcription, prenucleolar bodies (PNBs) are formed (Ochs et al., 1985; Azum-Gélade et al., 1994; Jiménez-Garcia et al., 1994; Dundr et al., 2000; Savino et al., 2001). The PNBs are discrete fibrogranular structures characterized by electron microscopy in animal and plant cells (Stevens, 1965; Ochs et al., 1985; Savino et al., 2001). PNB formation occurs on the chromosome surface during telophase and is under the control of cyclin-dependent kinases (Sirri et al., 2002). The PNBs also are formed during Xenopus development before complete nucleolar assembly, and they can be assembled in vitro in Xenopus egg extracts (Bell et al., 1992; Verheggen et al., 1998, 2001). Thus, PNB formation is a general phenomenon occurring during cell cycle and development. However, the role of this steady state along the recruitment pathway of the nucleolar processing complexes is presently unknown.
To investigate the role of PNBs in the establishment of nucleolar functions, we analyzed the dynamics and the possible interactions between processing proteins along the assembly pathway in living cells. The early rRNA processing machinery is cotranscriptionally associated with rRNAs, whereas the late rRNA processing machinery is involved in rRNA processing after termination of transcription. Here, we chose fibrillarin as representative protein of 90S pre-rRNA particles (early processing) and Nop52, Bop1, and B23 as representative proteins of 60s pre-rRNA particles (late processing) (Fatica and Tollervey, 2002; Fromont-Racine et al., 2003). The behavior of early and late rRNA processing proteins at the exit of mitosis was analyzed by 1) immunocytochemistry; 2) fast two-color four-dimensional (4D) imaging in living cells; and 3) time-lapse fluorescence resonance energy transfer (FRET) monitored by time domain fluorescence lifetime imaging microscopy (tdFLIM), a new, direct, and noninvasive method to follow protein–protein interactions in living cells. Based on data presented here, we propose that PNBs are preassembly platforms of rRNA processing complexes.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONACKNOWLEDGMENTSREFERENCES |
|---|
They were transfected at 60% confluence by using the Superfect reagent (QIAGEN, Valencia, CA). The following constructs were used: green fluorescent protein (GFP)-Nop52 and GFP-fibrillarin (Savino et al., 2001), GFP-B23 provided by Dr. S. Huang (Chen and Huang, 2001), GFP-Bop1 provided by Dr. D. Pestov (Pestov et al., 2001b), and Discosoma red fluorescence protein (DsRed)-Nop52 and DsRed-B23 inserted into pDsRed2-C1 (BD Biosciences Clontech, Palo Alto, CA). In all cases, GFP and DsRed were fused to the NH2 terminus of the proteins.
Several stably transformed cell lines were used: GFP-Nop52, GFP-fibrillarin, GFP-B23, GFP-Bop1, DsRed-B23, and DsRed-Nop52. Selection of the GFP-B23, GFP-Bop1, DsRed-B23, and DsRed-Nop52 stably transformed cells was carried out as described previously (Savino et al., 2001). Doubly transfected cells GFP-Nop52/DsRed-B23, GFP-fibrillarin/DsRed-B23, GFP-B23/DsRed-Nop52, GFP-Bop1/DsRed-B23, and DsRed-B23/GFP-Nop52 were generated from stably transformed cells transiently transfected with DsRed-B23, DsRed-Nop52, or GFP-Nop52. For FRET analysis, living adhesive cell lines permanently expressing the donor (GFP-Nop52, GFP-Bop1, or GFP-fibrillarin) were cultured on glass coverslips in Dulbecco's modified Eagle's medium, supplemented with 0.5% FCS, at 37°C in 5% CO2. The cells were transfected with DsRed-B23 1 d before measurements. The coverslips were mounted in a special holder allowing reconstruction of a Petri dish and placed on an inverted microscope. Measurements were carried out in culture medium without phenol red at 37°C in 5% CO2. A paraffin oil (Merck, Darmstadt, Germany) layer was deposited on top of the medium to avoid evaporation and to allow long-term observations at 37°C.
Antibodies
The following antibodies were used: a mouse monoclonal anti-fibrillarin, 72B9 (Reimer et al., 1987); a human serum directed against Nop52 (Savino et al., 1999); and a goat polyclonal anti-B23 (C19; Santa Cruz Biotechnology, Santa Cruz, CA). Anti-mouse secondary antibodies conjugated to fluorescein isothiocyanate (FITC) or to Cy5, anti-human antibodies conjugated to Cy5, and goat antibodies conjugated to Texas Red were from Jackson ImmunoResearch Laboratories (West Grove, PA). For immunolabeling, cells were treated as described previously (Savino et al., 2001).
Time-Lapse Microscopy
The cells were grown on glass coverslips, mounted in a Ludin observation chamber filled with complete medium supplemented with 10 mM HEPES, pH 7.4. The microscope and the chamber were kept at 37°C. A Leica DM IRB microscope, equipped with a piezoelectric translator (PIFOC; Physik Instrumente, Karlsruhe, Germany) placed at the base of a 100x PlanApo numerical aperture (NA) 1.4 objective and a 5-MHz Micromax 782Y interline charge-coupled device (CCD) camera (Roper Scientific, Evry Cedex, France) were used. To visualize two differently tagged proteins, rapid wavelength selection was achieved by a Shutter DG4 illuminator and wavelength changer. For imaging of GFP and DsRed, a dual narrow pass band FITC/tetramethylrhodamine B isothiocyanate (TRITC) filter block was used. In the DG-4 illuminator, short pass KP 500 and long pass LP 515 filters were mounted in positions 1 and 2, respectively. The acquisition software (MetaMorph; Universal Imaging, Downingtown, PA) was set to trigger rapid wavelength changes to acquire two images at each Z-step (0.3 µm). Imaging at full overlapped speed of the CCD device assured that the two fluorescent tags were recorded sequentially at maximum speed, without movement of the filters. The stacks were assembled after image deconvolution as described previously (Savino et al., 2001).
Fluorescence intensities in PNBs were quantified on the sum of three consecutive nondeconvoluted slices. The measurement on three slices was found necessary to include PNB movement. Measurements on regions of interest (ROI) were carried out in different subnuclear domains corresponding to diffuse or foci areas from telophase to early G1. Foci are defined as regions of local intensity greater than three times that of diffuse areas. The same ROI was used for GFP and DsRed fluorescence. The signals were quantified using the ImageJ software. The mean gray value, area, SD, minimum, and maximum gray value were recorded.
FRET Determination by tdFLIM Measurements
The apparatus used for FRET determination performs tdFLIM by the time- and space-correlated single-photon counting method and has been described previously (Emiliani et al., 2003). This technique directly gives the picosecond time-resolved fluorescence decay for every pixel by counting and sampling single emitted photons according to 1) the time delay between photon arrival and laser pulse (picosecond time scale, 4096 channels); 2) their x–y coordinate (256 x 256 pixel image), and 3) their absolute time. A titanium sapphire laser (Millennia 5W/Tsunami 3960-M3BB-UPG kit; Spectra-Physics, France) that delivers picosecond pulses was tuned at 960 nm to obtain a 480 nm excitation wavelength after frequency doubling. The repetition rate was 4 MHz after pulse-picker (3980-35; Spectra-Physics, Paris, France). The laser beam was expanded and inserted into an inverted epifluorescence microscope (Leica DMIRBE, Rueil-Malmaison, France) for wide field illumination. Green fluorescence decay images were taken using a Leica Plan-APOCHROMAT 100x1.3 NA oil objective, a dichroic beam splitter (505DRLP, Omega; Optophotonics, Eaubonne, France), an emission filter (535AF45, Omega; Optophotonics), and the quadrant-anode TSCSPC detector (QA; Europhoton, Berlin, Germany). The count rate was up to 50 kHz. The microscope stage was equipped with an incubator system for temperature and CO2 regulation (37°C, 5% CO2). TdFLIM images were obtained by mapping pixel by pixel the mean lifetime determined from a single exponential fit. For qualitative determination of the occurrence of FRET, the fluorescence decays corresponding to a nucleolus or PNB area were extracted from the acquisition matrix, and the decays of donor-tagged proteins in the presence of the acceptor-tagged B23 were compared with the control decays of donor-tagged proteins measured in the absence of acceptor. The experimental curves were further fitted with a Marquardt nonlinear least-square algorithm (Globals Unlimited software, University of Illinois, Urbana-Champaign, IL) by using a Gaussian distribution of lifetimes as theoretical model. Time-lapse FRET images were obtained from the Gaussian distribution analysis of the fluorescence decays associated with each region of interest and at different elapsed times of acquisition. A reduction in the center of the fluorescence lifetime distribution of GFP-tagged protein superior or equal to 0.200 ns was considered as occurrence of FRET.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONACKNOWLEDGMENTSREFERENCES |
|---|
), by the presence of the proteins at the surface of isolated chromosomes (Gautier et al., 1992b), and by localization of the proteins in electron microscopy (Gautier et al., 1992a). During telophase and early G1, the rRNA processing proteins form PNBs at the chromosome periphery and are progressively recruited to the site of transcription. As already described, early processing proteins (fibrillarin) are recruited first, whereas the majority of late processing proteins are in PNBs (Savino et al., 1999). It is noticeable that Bop1 and B23 are in the same PNBs (Figure 2), as well as Nop52 (our unpublished data). Fibrillarin is also in PNBs before recruitment to the nucleolus during a short period (Savino et al., 2001). We cannot exclude the possibility that early and late processing proteins are in the same PNBs during this short period (
10 min) of telophase as visible in Figure 2.
|
|
PNBs Contain Both Early and Late rRNA Processing Proteins in Telophase
We assessed the relative dynamics of early and late rRNA processing proteins at the time of PNB formation. Using cells coexpressing GFP-fibrillarin and DsRed-B23, we analyzed in the same living cell the kinetics of translocation of the two proteins by rapid two-color 4D imaging (volume + time for both proteins). Mitotic cells expressing a similar and low level of both tagged proteins were selected, eliminating cells with high expression. We assume that GFP-fibrillarin is functional because the permanent cell line was selected on the basis of identical distribution for endogenous and GFP-fibrillarin, of normal cell cycle, and of the level of GFP-fibrillarin (Savino et al., 2001). In all the cells (n = 24) studied, once near the poles, one or two min after the onset of telophase, numerous bright fluorescent foci containing both GFP-fibrillarin and DsRed-B23 became visible almost simultaneously (Figure 3 and Video 2). In particular, in corresponding focal planes, the same foci contained GFP-fibrillarin and DsRed-B23 and looked orange in the color merge (Figure 3, arrow, and Video 2). To confirm this observation, the dynamics of this colocalization was measured. GFP and DsRed were quantified in the same foci and compared with dispersed proteins in the same nucleus (as described in Materials and Methods). The identification of foci corresponding to PNBs versus incipient nucleoli was based on the gradual growth of the structures with time; incipient nucleoli increased in size and intensity contrary to PNBs (see GFP-fibrillarin, Figure 3). Also, several tens of PNBs are formed, but only a maximum of six incipient nucleoli in this cell line (Roussel et al., 1996). Starting in telophase, 12 PNBs were analyzed in four different cells for periods lasting 20 min (Figure 4). The relative amount of B23 in PNBs was 5 to 6 times that of the dispersed proteins (Figure 4, compare continuous and dotted red curves). Interestingly, the amount of fibrillarin in these PNBs was 3 to 4 times that of dispersed proteins for 10 min (Figure 4, compare continuous and dotted green curves). After this time, fibrillarin was released, whereas B23 was still present in PNBs, either decreasing or remaining at high levels (Figure 4A, 19', and graphs B and C). This clearly illustrates the presence of the two types of nucleolar processing proteins in the same PNBs and suggests differential sorting of these proteins.
|
|
Accordingly, in all the cells studied, the timing of recruitment of these two proteins into incipient nucleoli was different (Figures 2 and 3 and Video 2). In the example shown, 10 min after the onset of telophase, GFP-fibrillarin was seen in incipient nucleoli, whereas in the same nucleoli, DsRed-B23 was detected only 20 min later (Figure 3 and Video 2).
Similar Dynamics of Nop52 and B23
The dynamics of GFP-Nop52 and DsRed-B23 were analyzed simultaneously in the same living cells. Both proteins participate in processing of the internal transcribed spacer 2 of rRNAs (Savkur and Olson, 1998; Savino et al., 1999). We found similar dynamics and flows of both proteins within the same PNBs when simultaneously observed by two-color 4D microscopy (Figure 5A and Video 3). The dynamics of this colocalization was quantified in the same PNBs as described for fibrillarin and B23. Starting in telophase, eight PNBs were analyzed in four different cells for periods lasting 25 min. The relative amounts of Nop52 and B23 in PNBs were very similar (Figure 5C, compare red and green curves). Interestingly, the amounts of both proteins varied simultaneously in particular when decreasing (Figure 5C, arrows). This could indicate that these late processing proteins are released from PNBs as complexes and could interact along the pathway and be recruited together into nucleoli (see Figure 5 legend for details). This prompted us to investigate the interactions of these proteins from the chromosome periphery to PNBs and nucleoli.
|
; Tramier et al., 2002) was not detected in our conditions. Consequently, the reduction of the fluorescence lifetime of GFP-Nop52 measured in the presence of DsRed-B23 does not arise from a DsRed-associated artifact. In addition, it was not possible to detect FRET between two late processing proteins, GFP-Bop1 and DsRed-B23 (our unpublished data), indicating that colocalization in the nucleolar granular component is not sufficient to record a positive FRET signal. Together, these results indicate that in nucleoli, FRET occurs between GFP-Nop52 and DsRed-B23.
|
The kinetics of decay of nucleolus-associated fluorescence in the presence or absence of DsRed-B23, of, respectively, GFP-Nop52 (Figure 6c) and GFP-fibrillarin (Figure 6c') was analyzed by Gaussian distributions of the fluorescence lifetime (Figure 6, d and d', respectively). Each bell-shaped curve represents the lifetime distributions of GFP-Nop52 fluorescence (Figure 6d) and of GFP-fibrillarin fluorescence (Figure 6d') within a given single nucleolus, measured in the absence (green curves) and in the presence (red curves) of the acceptor (DsRed-B23). The entire fluorescence lifetime distribution of GFP-Nop52 measured in the presence of DsRed-B23 in several nucleoli (n = 35) was shifted toward short fluorescence lifetime values compared with the distribution measured in the absence of DsRed-B23 (n = 53) (Figure 6d). This was accompanied by a concomitant broadening of the fluorescence lifetime distribution (Figure 6d). In contrast, the fluorescence lifetime distributions of GFP-fibrillarin measured in the absence (n = 22) and presence (n = 21) of DsRed-B23 display complete overlap with each other (Figure 6d'). The shift of the center of the fluorescence lifetime distribution in nucleoli of GFP-Nop52/DsRed-B23 cells (from 2.33 ± 0.05 ns in control GFP-Nop52 cells to 2.04 ± 0.06 ns) makes it possible to calculate a mean FRET efficiency of 12.4% between GFP-Nop52 and DsRed-B23 in nucleoli of living cells (from the expression: E = 1 – 
DA/D, with E being the FRET efficiency, DA the fluorescence life-time of the donor in the presence of the acceptor, and D the fluorescence lifetime of the donor alone). This is to be compared with the absence of shift between the center of the fluorescence lifetime distribution in nucleoli of GFP-fibrillarin/DsRed-B23 cells (2.36 ± 0.04 ns) and that in nucleoli of control GFP-fibrillarin cells (2.34 ± 0.04 ns). The broadening of the lifetime distribution in nucleoli of GFP-Nop52/DsRed-B23 cells provides evidence of an increasing heterogeneity of GFP lifetime due to variable situations of GFP-Nop52/DsRed-B23 proximity/orientation in the same nucleolus.
Late rRNA Processing Proteins Already Interact in PNBs
Because it is possible to detect FRET between B23 and Nop52 in nucleoli, we decided to track FRET during the recruitment of these proteins into nucleoli from anaphase to early G1. The principles of the analysis are presented in Figure 7, i.e., acquisition of GFP fluorescence decay images in living cells, then manual drawing of the regions of interest around the GFP signals and analysis of the FRET in these regions throughout recording. Time-lapse tdFLIM-FRET measurements were carried out by acquiring fluorescence decay images of the GFP donor in GFP-Nop52/DsRed-B23 along the PNB pathway from late anaphase to the early G1 (Figure 8, a–f); only one of the two daughter cells is presented in Figure 8. We carried out FRET analysis for 35 PNBs and nucleoli in the same living cell. Analysis of the experimental data showed that FRET was never detected during anaphase at the periphery of the chromosomes (Figure 8a'), whereas FRET was registered in 20% PNBs at the beginning of telophase (Figure 8b'), in 40% at the end of telophase (Figure 8, c' and d'), and in 55% in early G1 (Figure 8, e' and f'). Thus, interaction between GFP-Nop52 and DsRed-B23 was established progressively in PNBs, as the number of PNBs exhibiting FRET increased. Such data indicate that Nop52 and B23 did not interact until they were recruited in PNBs. It is noteworthy that a given PNB can alternatively present FRET or not present FRET (for example, see PNB 16, Figure 8).
|
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONACKNOWLEDGMENTSREFERENCES |
|---|
; Ochs et al., 1985; Jiménez-Garcia et al., 1994; Savino et al., 1999, 2001; Dundr et al., 2000). This is a general phenomenon that occurs in most cells, but the role of these large complexes in the delivery of the rRNA processing machinery is presently unclear.
The formation of PNBs depends on the inactivation of the cyclin-dependent kinase (CDK) 1-cyclin B, and recruitment in the nucleoli of early and late processing proteins is differentially regulated (Sirri et al., 2002). The processing proteins cotranscriptionally associated with rRNA such as fibrillarin are recruited first, followed by proteins involved in late steps of processing such as B23 and Nop52 (Savino et al., 2001; Leung et al., 2004). Accordingly, in telophase no granular component is visible in incipient nucleoli (Hernandez-Verdun et al., 1980). PNB formation could contribute to the temporal order of recruitment of the processing complexes. This temporal order could be established either by different PNBs with different lifetimes or by differential sorting from the same PNBs. The former possibility has been the preferred one so far (Ochs et al., 1985; Jiménez-Garcia et al., 1994; Savino et al., 2001). In the present study, comparing the kinetics of early and late processing proteins in the same cells, we obtained evidence that initially both early and late processing proteins are concentrated in the same PNBs from which they are differentially sorted. In retrospect, given the short time window of the events described here (10 min), it was important to compare the kinetics in the same cells by using the rapid two-color 4D microscopy approach, and this explains why this mechanism was not reported previously.
An important question concerns the timing of the formation of the rRNA processing protein complexes. Here, we investigated during nucleolar assembly, the interaction of Nop52 and B23, proteins involved in the processing of 60S ribosome particles. The choice of these proteins was dictated by their function in the internal transcribed spacer 2 and because immunoprecipitation indicated that they are in the same complexes (our unpublished data). In yeast, the recent advent of proteomic analyses demonstrates that rRNA processing implies large protein complexes associated sequentially to nascent ribosomal particles. These complexes are specific of each processing step and are different for 40 and 60S ribosomal subunits (Dragon et al., 2002; Fatica and Tollervey, 2002; Fromont-Racine et al., 2003; Horsey et al., 2004). In mammals, the characterization of these rRNA processing complexes lags far behind that in yeast. However, it is possible to take advantage of the high conservation of some rRNA processing proteins between yeast and humans to determine the roles of the human proteins. Rrp1p is the yeast homologue of the human Nop52 (Savino et al., 1999; Horsey et al., 2004). A proteomic analysis of the Rrp1p partners has demonstrated that the preribosome particles contain at least 28 nonribosomal proteins necessary for the production of 60S (Horsey et al., 2004). Interestingly, Erb1p, the yeast homologue of the human Bop1 (Pestov et al., 2001a), was detected in the complex. No yeast homologue of B23 has been identified, but this protein is known to be involved in the processing of the ITS2 (Savkur and Olson, 1998) as Nop52 and Bop1.
To analyze protein interactions in living cells, energy transfer and consequently the distance between GFP and DsRed of tagged processing proteins was determined using tdFLIM. Monitoring FRET by tdFLIM has the advantage of being independent of chromophore concentration and lightpath length (Bastiaens and Squire, 1999; Selvin, 2000), as opposed to steady-state fluorescence intensity-based measurements. The use of a time-correlated single-photon counting method to perform tdFLIM (Emiliani et al., 2003; Tramier et al., 2003) provides the additional advantage of allowing precise determination of short fluorescence lifetimes and of detecting small variations in these lifetimes due to the excellent time resolution (100 ps) of the fluorescence decay kinetics. Moreover, this technique is minimally invasive because it requires only a very low level of laser excitation intensity (Gautier et al., 2001; Tramier et al., 2002). It is therefore very well suited to carry out time-lapse FRET imaging microscopy in living cells during extended periods.
In contrast to other GFP-tagged proteins for which fluorescence lifetimes have already been analyzed in living cells (Gautier et al., 2001; Tramier et al., 2002, 2003), the fluorescence decays of nucleolar GFP-tagged proteins GFP-fibrillarin, GFP-Nop52, GFP-Bop1, and GFP-B23 could not be fitted with a single fluorescence lifetime model in the absence of the acceptor (our unpublished data). This particularity is not understood and might be due to various conformational constraints on GFP when embedded in huge macromolecular complexes such as nucleoli and PNBs. The fact that the fluorescence decay of the donor (GFP-Nop52) was not monoexponential prevented a quantitative analysis of FRET (determination of the intrinsic FRET efficiency and of the ratio of bound and unbound donor). Such an analysis indeed requires the existence of a single fluorescence lifetime of GFP in the absence of acceptor that would allow the determination of both FRET parameters (Tramier et al., 2002; Emiliani et al., 2003). Here, the presence of FRET between GFP-Nop52 and DsRed-B23 was deduced from the shift between the center of the GFP-Nop52 lifetime distribution in the nucleoli and PNBs of GFP-Nop52/DsRed-B23 cells and of GFP-Nop52–containing cells. The associated broadening of the lifetime distribution suggests the existence of a distribution of the true FRET efficiency between GFP-Nop52 and DsRed-B23 (conformational changes within the nucleus, i.e., distance/orientation changes between donor and acceptor), and/or the existence of various relative amounts of bound and unbound GFP-Nop52 within a single nucleolus or PNB. Time-lapse tdFLIM experiments showed that the FRET signal is transient within a single PNB. This could arise from either transient interactions between GFP-Nop52 and DsRed-B23 within a single PNB or stable interactions but release of the GFP-Nop52/DsRed-B23 complexes toward the nucleoli. Both explanations reveal unanticipated dynamics within the PNB compartment. However, the fact that the steady-state intensity of GFP within a single PNB also decreased during nucleoli reconstruction suggests that the release of GFP-Nop52 and DsRed-B23 toward the nucleoli most likely occurs in the form of stable complexes. This hypothesis also supported by the simultaneous periodic decrease of both proteins from PNBs observed in two-color 4D imaging (Figure 5C).
The formation of PNBs occurs during telophase. Given the dynamic nature of many of the nucleolar proteins composing these bodies (Phair and Misteli, 2000; Huang, 2002), it is surprising that these structures exist at precisely this period of the cell cycle. As proposed for other nuclear bodies (Janicki and Spector, 2003), high-affinity binding sites could induce the formation of the PNBs. This is exactly what was found in the present study. During translocation of the rRNA processing proteins, interactions between GFP-Nop52 and DsRed-B23 were detected in PNBs and not at the chromosome periphery where these two proteins colocalize. Hence, the capacity of these two proteins to interact is modified at the telophase/early G1 transition and is initiated in PNBs. Such interactions between Nop52 and B23 occur in PNBs even before their recruitment in the incipient nucleoli (Figure 9). The presence of rRNAs as well as of small nucleolar RNAs has been demonstrated in PNBs (Jiménez-Garcia et al., 1994; Verheggen et al., 1998; Dousset et al., 2000). The interaction of B23 with rRNAs depends on phosphorylation of the protein by CDK (Okuwaki et al., 2002). In particular, the RNA binding activity of B23 is disrupted by CDK1-cyclin B during mitosis. Thus, the presence of rRNAs could contribute to the assembly of the rRNA processing complexes in PNBs when CDK1-cyclin B is inactivated, an event that normally occurs at the end of mitosis. The release of late processing proteins (B23 and Nop52) from PNBs (not the early ones) seems to be regulated by a presently uncharacterized CDK (Sirri et al., 2002). However, it is impossible to predict what inhibits this PNB release because the interactions depending on phosphorylation of this CDK are not characterized. One possible explanation could be that the late processing machinery assembled with rRNAs in PNBs is activated by this unknown CDK, thus becoming able to process rRNA and consequently be released from PNBs. This model could explain why formation of PNBs occurs only during telophase/early G1 transition and why the recruitment of late processing proteins takes longer than that of early ones. However, uncovering the mechanisms that govern the recruitment of the nucleolar processing machinery will require further investigation of the interactions occurring in PNB complexes and of the transit between PNBs and the nucleolus.
|
In conclusion, we observed that early and late rRNA processing proteins pass through the same PNBs, that differential sorting from the same PNBs ensures differential nucleolar recruitment, and that interactions between protein partners occur in PNBs. We therefore propose that PNBs function as assembly platforms of rRNA processing complexes.
|
ACKNOWLEDGMENTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONACKNOWLEDGMENTSREFERENCES |
|---|
|
Footnotes |
|---|
Abbreviations used: DsRed, Discosoma red fluorescence protein; FRET, fluorescence resonance energy transfer; GFP, green fluorescence protein; PNB, prenucleolar body; ROI, regions of interest; tdFLIM, time domain fluorescence lifetime imaging microscopy.
The online version of this article contains supplemental material at MBC Online (http://www.molbiolcell.org). 
Address correspondence to: Danièle Hernandez-Verdun (dhernand{at}ccr.jussieu.fr).
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONACKNOWLEDGMENTSREFERENCES |
|---|
Bastiaens, P. I., and Squire, A. ((1999). ). Fluorescence lifetime imaging microscopy: spatial resolution of biochemical processes in the cell. Trends Cell Biol. 9, , 48–52.[CrossRef][Medline]
Bell, P., Dabauvalle, M. C., and Scheer, U. ((1992). ). In vitro assembly of prenucleolar bodies in Xenopus egg extract. J. Cell Biol. 118, , 1297–1304.
Bubulya, P. A., Prasanth, K. V., Deerinck, T. J., Gerlich, D., Beaudouin, J., Ellisman, M. H., Ellenberg, J., and Spector, D. L. ((2004). ). Hypophosphorylated SR splicing factors transiently localize around active nucleolar organizing regions in telophase daughter nuclei. J. Cell Biol. 167, , 51–63.
Chen, D., and Huang, S. ((2001). ). Nucleolar components involved in ribosome biogenesis cycle between the nucleolus and nucleoplasm in interphase cells. J. Cell Biol. 153, , 169–176.
Chubb, J. R., Boyle, S., Perry, P., and Bickmore, W. A. ((2002). ). Chromatin motion is constrained by association with nuclear compartments in human cells. Curr. Biol. 12, , 439–445.[CrossRef][Medline]
Cotlet, M., Hofkens, J., Habuchi, S., Dirix, G., Van Guyse, M., Michiels, J., Vanderleyden, J., and De Schryver, F. C. ((2001). ). Identification of different emitting species in the red fluorescent protein DsRed by means of ensemble and single-molecule spectroscopy. Proc. Natl. Acad. Sci. USA 98, , 14398–14403.
Dousset, T., Wang, C., Verheggen, C., Chen, D., Hernandez-Verdun, D., and Huang, S. ((2000). ). Initiation of nucleolar assembly is independent of RNA polymerase I transcription. Mol. Biol. Cell 11, , 2705–2717.
Dragon, F., et al. ((2002). ). A large nucleolar U3 ribonucleoprotein required for 18S ribosomal RNA biogenesis. Nature 417, , 967–970.[CrossRef][Medline]
Dundr, M., Misteli, T., and Olson, M.O.J. ((2000). ). The dynamics of postmitotic reassembly of the nucleolus. J. Cell Biol. 150, , 433–446.
Emiliani, V., Sanvitto, D., Tramier, M., Piolot, T., Petrasek, Z., Kemnitz, K., Durieux, C., and Coppey-Moisan, M. ((2003). ). Low-intensity two-dimensional imaging of fluorescence lifetimes in living cells. Appl. Phys. Lett. 83, , 2471–2473.[CrossRef]
Fatica, A., and Tollervey, D. ((2002). ). Making ribosomes. Curr. Opin. Cell Biol. 14, , 313–318.[CrossRef][Medline]
Fromont-Racine, M., Senger, B., Saveanu, C., and Fasiolo, F. ((2003). ). Ribosome assembly in eukaryotes. Gene 313, , 17–42.[CrossRef][Medline]
Gall, J. G. ((2000). ). Cajal bodies: the first 100 years. Annu. Rev. Cell Dev. Biol. 16, , 273–300.[CrossRef][Medline]
Gautier, I., Tramier, M., Durieux, C., Coppey, J., Pansu, R. B., Nicolas, J. C., Kemnitz, K., and Coppey-Moisan, M. ((2001). ). Homo-FRET microscopy in living cells to measure monomer-dimer transition of GFT-tagged proteins. Biophys. J. 80, , 3000–3008.
Gautier, T., Dauphin-Villemant, C., André, C., Masson, C., Arnoult, J., and Hernandez-Verdun, D. ((1992a). ). Identification and characterization of a new set of nucleolar ribonucleoproteins which line the chromosomes during mitosis. Exp. Cell Res. 200, , 5–15.[CrossRef][Medline]
Gautier, T., Robert-Nicoud, M., Guilly, M.-N., and Hernandez-Verdun, D. ((1992b). ). Relocation of nucleolar proteins around chromosomes at mitosis - a study by confocal laser scanning microscopy. J. Cell Sci. 102, , 729–737.
Hadjiolov, A. A. ((1985). ). The Nucleolus and Ribosome Biogenesis, Vienna, Springer.
Hernandez-Verdun, D., Bourgeois, C. A., and Bouteille, M. ((1980). ). Simultaneous nucleologenesis in daughter cells during late telophase. Biol. Cell 37, , 1–4.
Hernandez-Verdun, D., Roussel, P., and Gébranne-Younès, J. ((2002). ). Emerging concepts of nucleolar assembly. J. Cell Sci. 115, , 2265–2270.
Horsey, E. W., Jakovljevic, J., Miles, T. D., Harnpicharnchai, P., and Woolford, J. L., Jr. ((2004). ). Role of the yeast Rrp1 protein in the dynamics of pre-ribosome maturation. RNA 10, , 813–827.
Huang, S. ((2002). ). Building an efficient factory: where is pre-rRNA synthesized in the nucleolus? J. Cell Biol. 157, , 739–741.
Janicki, S. M., and Spector, D. L. ((2003). ). Nuclear choreography: interpretations from living cells. Curr. Opin. Cell Biol. 15, , 149–157.[CrossRef][Medline]
Jiménez-Garcia, L. F., Segura-Valdez, M. L., Ochs, R. L., Rothblum, L. I., Hannan, R., and Spector, D. L. ((1994). ). Nucleologenesis: U3 snRNA-containing prenucleolar bodies move to sites of active pre-rRNA transcription after mitosis. Mol. Biol. Cell 5, , 955–966.[Abstract]
Lamond, A. I., and Earnshaw, W. C. ((1998). ). Structure and function in the nucleus. Science 280, , 547–553.
Lamond, A. I., and Spector, D. L. ((2003). ). Nuclear speckles: a model for nuclear organelles. Nat. Rev. Mol. Cell. Biol. 4, , 605–612.[CrossRef][Medline]
Leung, A. K., Andersen, J. S., Mann, M., and Lamond, A. I. ((2003). ). Bioinformatic analysis of the nucleolus. Biochem. J. 376, , 553–569.[CrossRef][Medline]
Leung, A. K., Gerlich, D., Miller, G., Lyon, C., Lam, Y. W., Lleres, D., Daigle, N., Zomerdijk, J., Ellenberg, J., and Lamond, A. I. ((2004). ). Quantitative kinetic analysis of nucleolar breakdown and reassembly during mitosis in live human cells. J. Cell Biol. 166, , 787–800.
Matera, A. G. ((1999). ). Nuclear bodies: multifaceted subdomains of the interchromatin space. Trends Cell Biol. 9, , 302–309.[CrossRef][Medline]
Mélèse, T., and Xue, Z. ((1995). ). The nucleolus: an organelle formed by the act of building a ribosome. Curr. Opin. Cell Biol. 7, , 319–324.
Misteli, T. ((2000). ). Cell biology of transcription and pre-mRNA splicing: nuclear architecture meets nuclear function. J. Cell Sci. 113, , 1841–1849.[Abstract]
Ochs, R. L., Lischwe, M. A., Shen, E., Caroll, R. E., and Busch, H. ((1985). ). Nucleologenesis: composition and fate of prenucleolar bodies. Chromosoma 92, , 330–336.[CrossRef][Medline]
Okuwaki, M., Tsujimoto, M., and Nagata, K. ((2002). ). The RNA binding activity of a ribosome biogenesis factor, nucleophosmin/B23, is modulated by phosphorylation with a cell cycle-dependent kinase and by association with its subtype. Mol. Biol. Cell 13, , 2016–2030.
Olson, M.O.J., Dundr, M., and Szebeni, A. ((2000). ). The nucleolus: an old factory with unexpected capabilities. Trends Cell Biol. 10, , 189–196.[CrossRef][Medline]
Pederson, T. ((1998). ). The plurifunctional nucleolus. Nucleic Acids Res. 26, , 3871–3876.
Pestov, D. G., Stockelman, M. G., Strezoska, Z., and Lau, L. F. ((2001a). ). ERB1, the yeast homolog of mammalian Bop1, is an essential gene required for maturation of the 25S and 5.8S ribosomal RNAs. Nucleic Acids Res. 29, , 3621–3630.
Pestov, D. G., Strezoska, Z., and Lau, L. F. ((2001b). ). Evidence of p53-dependent cross-talk between ribosome biogenesis and cell cycle: effects of nucleolar protein Bop1 on G1/S transition. Mol. Cell. Biol. 21, , 4246–4255.
Phair, R. D., and Misteli, T. ((2000). ). High mobility of proteins in the mammalian cell nucleus. Nature 404, , 604–609.[CrossRef][Medline]
Prasanth, K. V., Sacco-Bubulya, P. A., Prasanth, S. G., and Spector, D. L. ((2003). ). Sequential entry of components of gene expression machinery into daughter nuclei. Mol. Biol. Cell 14, , 1043–1057.
Reimer, G., Pollard, K. M., Pennig, C. A., Ochs, R. L., Lischwe, M. A., Busch, H., and Tan, E. M. ((1987). ). Monoclonal autoantibody from NZB/NZW F1 mouse and some human scleroderma sera target a Mr 34000 nucleolar protein of the U3-ribonucleoprotein particle. Arthritis Rheum. 30, , 793–800.[Medline]
Roussel, P., André, C., Comai, L., and Hernandez-Verdun, D. ((1996). ). The rDNA transcription machinery is assembled during mitosis in active NORs and absent in inactive NORs. J. Cell Biol. 133, , 235–246.
Savino, T. M., Bastos, R., Jansen, E., and Hernandez-Verdun, D. ((1999). ). The nucleolar antigen Nop52, the human homologue of the yeast ribosomal RNA processing RRP1, is recruited at late stages of nucleologenesis. J. Cell Sci. 112, , 1889–1900.[Abstract]
Savino, T. M., Gébrane-Younès, J., De Mey, J., Sibarita, J.-B., and Hernandez-Verdun, D. ((2001). ). Nucleolar assembly of the rRNA processing machinery in living cells. J. Cell Biol. 153, , 1097–1110.
Savkur, R. S., and Olson, M.O.J. ((1998). ). Preferential cleavage in pre-ribosomal RNA by protein B23 endoribonuclease. Nucleic Acids Res. 26, , 4508–4515.
Selvin, P. R. ((2000). ). The renaissance of fluorescence resonance energy transfer. Nat. Struct. Biol. 7, , 730–734.[CrossRef][Medline]
Sirri, V., Hernandez-Verdun, D., and Roussel, P. ((2002). ). Cyclin-dependent kinases govern formation and maintenance of the nucleolus. J. Cell Biol. 156, , 969–981.
Spector, D. L. ((1993). ). Macromolecular domains within the cell nucleus. Annu. Rev. Cell Biol. 9, , 265–315.[CrossRef][Medline]
Spector, D. L. ((2001). ). Nuclear domain. J. Cell Sci. 114, , 2891–2893.
Stevens, B. ((1965). ). The fine structure of the nucleolus during mitosis in the grasshopper neuroblast cell. J. Cell Biol. 24, , 349–368.
Tramier, M., Gautier, I., Piolot, T., Ravalet, S., Kemnitz, K., Coppey, J., Durieux, C., Mignotte, V., and Coppey-Moisan, M. ((2002). ). Picosecond-hetero-FRET microscopy to probe protein-protein interactions in live cells. Biophys. J. 83, , 3570–3577.
Tramier, M., Piolot, T., Gautier, I., Mignotte, V., Coppey, J., Kemnitz, K., Durieux, C., and Coppey-Moisan, M. ((2003). ). Homo-FRET versus hetero-FRET to probe homodimers in living cells. Methods Enzymol. 360, , 580–597.[Medline]
Verheggen, C., Le Panse, S., Almouzni, G., and Hernandez-Verdun, D. ((1998). ). Presence of pre-rRNAs before activation of polymerase I transcription in the building process of nucleoli during early development of Xenopus laevis. J. Cell Biol. 142, , 1167–1180.
Verheggen, C., Le Panse, S., Almouzni, G., and Hernandez-Verdun, D. ((2001). ). Maintenance of nucleolar machineries and pre-rRNAs in remnant nucleolus of erythrocyte nuclei and remodeling in Xenopus egg extracts. Exp. Cell Res. 269, , 23–34.[CrossRef][Medline]
Zatsepina, O. V., Todorov, I. T., Philipova, R. N., Krachmarov, C. P., Trendelenburg, M. F., and Jordan, E. G. ((1997). ). Cell cycle-dependent translocations of a major nucleolar phosphoprotein, B23, and some characteristics of its variants. Eur. J. Cell Biol. 73, , 58–70.[Medline]
This article has been cited by other articles:
|
|
 |
|
  H. Ma and T. Pederson Nucleophosmin Is a Binding Partner of Nucleostemin in Human Osteosarcoma Cells Mol. Biol. Cell, July 1, 2008; 19(7): 2870 - 2875. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. Mijatovi |
< 560 nm) in permanent cell lines expressing GFP-Nop52 (a–d) or GFP-fibrillarin (a'–d'), alone (a, a', and green curves) or in the presence of the DsRed acceptor after transfection with DsRed-B23 (b, b', and red curves). The tdFLIM images were obtained by analyzing pixel by pixel the fluorescence decays with a single lifetime and are displayed as fluorescence lifetime pseudocolored maps (a, a', b, and b'). The lifetime between 2.45 and 1.85 ns is indicated by colors presented in the scale. The nucleolus-associated fluorescence decay of GFP-tagged proteins is visible and compared in c and c' to donor alone (green curves, GFP-Nop52 and GFP-fibrillarin, respectively) and to donor with acceptor (red curves, DsRed-B23). Fits of these fluorescence decays were carried out using a Gaussian distribution lifetime model, and the complete results are plotted in d for Nop52GFP alone (green curves, n = 53) and in the presence of B23DsRed (red curves, n = 35), and in d' for GFP-fibrillarin alone (green curves, n = 22) and in the presence of DsRed-B23 (red curves, n = 21). Bar, 10 µm.
