![[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. 15, Issue 4, 1816-1832, April 2004
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
* Division of Cell Biology, German Cancer Research Center, D-69120 Heidelberg, Germany;
Protein Analysis Facility, German Cancer Research Center, D-69120 Heidelberg, Germany; and
Cytometry Facility, German Cancer Research Center, D-69120 Heidelberg, Germany
Submitted August 25, 2003;
Revised December 17, 2003;
Accepted January 8, 2004
Monitoring Editor: Joseph Gall
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONACKNOWLEDGMENTSREFERENCES |
|---|
, and PCNA, respectively, have further shown that the staining pattern of NO66 overlaps with certain clusters of late replicating chromatin. Biochemical experiments have revealed that protein NO66 cofractionates with large preribosomal particles but is absent from cytoplasmic ribosomes. We propose that in addition to its role in ribosome biogenesis protein NO66 has functions in the replication or remodeling of certain heterochromatic regions. |
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONACKNOWLEDGMENTSREFERENCES |
|---|
; Grummt, 2003). More recent evidence, however, has shown that the nucleolus is also involved in the assembly of various other kinds of ribonucleoprotein particles, the modification of small RNAs, the control of the cell cycle, the sequestration of regulatory molecules, and nuclear export processes (reviewed by Pederson, 1998; Olson et al., 2002; Gerbi et al., 2003). The discovery of novel functional importance of the nucleolus was paralleled by two recent proteomic analyses of human nucleoli (Andersen et al., 2002; Scherl et al., 2002), in which a total of 
350 different proteins have been identified, adding further support to the concept of the plurifunctional nature of nucleoli.
Morphologically, the nucleolus is characterized by the presence of three major structural components defined by electron microscopy: The internal fibrillar center (FC) is surrounded by the dense fibrillar component (DFC) and the granular component (GC), constituting the bulk of an active nucleolus. Localization studies using specific antibodies as well as hybridization probes have disclosed that the vectorial process of ribosome synthesis can be correlated with distinct nucleolar substructures, i.e., nascent preribosomes move from the DFC region to the peripherally located GC (e.g., Thiry et al., 2000; Huang, 2002). In addition, a nucleolus-specific karyoskeletal element has been demonstrated in the nucleolar cortex of amphibian oocytes (Franke et al., 1981; Kneissel et al., 2001).
Extended immunolocalization studies of nuclear proteins and, in particular, live-cell imaging have disclosed that nuclear processes rely on a constant flow of molecules between nuclear subcompartments (reviewed by Carmo-Fonseca, 2002; Leung and Lamond, 2003). Consequently, certain nuclear proteins may not be restricted to one nuclear substructure alone, but may also occur—at least transiently or in special phases—in other nuclear substructures. Indeed, a number of nucleolar proteins, such as fibrillarin, Nopp140, and NAP57, have been also found in Cajal bodies (Ochs et al., 1985; Meier and Blobel, 1990, 1994; Raska et al., 1991), and the survival of motor neuron (SMN) protein as well as its interacting proteins have been localized to "gems" and nucleoli (Charroux et al., 2000; Wehner et al., 2002). Under certain conditions proteins normally found in promyelotic leukemia (PML) bodies or "paraspeckles" can move to the nucleolus (Lin and Shih, 2002; Fox et al., 2002), and protein Ki-67, a widely used tumor marker, localizes to both nucleoli and heterochromatic regions (Starborg et al., 1996; Bridger et al., 1998).
The number of nucleoli per nucleus can vary greatly, from one or a few located at chromosomal nucleolar organizers, to more than thousand extrachromosomal nucleoli in certain amphibian oocytes (Hadjiolov, 1985). The presence of a high copy number of rRNA genes and the absence of nonribosomal DNA make the Xenopus oocyte nucleoli a particularly valuable model to analyze nucleolar proteins and their functions. This prompted us to improve the purification of amplified nucleoli from Xenopus oocyte nuclei by fluorescence-activated particle sorting, originally described by Franke et al. (1981). The resulting "pure" nucleolar fraction was subsequently analyzed with protein chemical and immunolocalization methods.
Here, we report the identification and molecular characterization of a so far unknown nucleolar protein of 66 kDa from Xenopus laevis, hence termed protein NO66. This protein, highly conserved from Caenorhabditis elegans to man, displays an unusual distribution: Besides a strong accumulation in nucleoli, it is also localized in distinct intranuclear bodies that represent clusters of late replicating chromatin. In biochemical studies we further show that in the nucleolus NO66 is enriched in large, most likely preribosomal particles, and on immunoprecipitation it copurifies with a number of well-characterized nucleolar constituents. We conclude that we have identified a dual location intranuclear protein that on the one hand plays a role in ribosome biogenesis and on the other participates in the replication or silencing of certain heterochromatic regions.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONACKNOWLEDGMENTSREFERENCES |
|---|
Cell culture lines used in this study included vulvar squamous carcinoma line A-431, astrocytoma derived glioma line U333 CG/343 MG, spontaneously transformed keratinocyte line HaCaT, cervical adenocarcinoma line HeLa, breast adenocarcinoma line MCF-7, primary liver carcinoma line PLC, SV40-transformed fibroblast line SV80 (all of human origin), mouse embryonic fibroblast line 3T3, rat smooth muscle cell line RV, bovine mammary gland-derived cell line BMGE+H, canine kidney cell line MDCK, potoroo kidney cell line PtK2, chicken embryonic fibroblast line CEF, rainbow trout gonadal cell line RTG-2, and X. laevis kidney epithelium line XLKE-A6. For sources of all cell lines see American Type Culture Collection (ATCC, Manassas, VA) and previous reports from this laboratory (Franke et al., 1979; Schmid et al., 1983; Cordes et al., 1996). Primary human umbilical vein endothelial cells (HUVEC) were prepared as described by Peitsch et al. (1999).
HeLa cells were synchronized at the G1/S-phase border by a double-thymidine block according to standard protocols. Cells were released by adding fresh culture medium and their progression toward the cell cycle was analyzed at different time points by flow cytometry and immunofluorescence microscopy.
Large-scale Isolation and Fractionation of X. laevis Oocyte Nuclei and Preparation of Egg Extracts
Nuclei from mature X. laevis oocytes (stages IV-VI; Dumont, 1972) were obtained by mass isolation using the procedure of Scalenghe et al. (1978), with modifications of Kleinschmidt and Franke (1982). Subsequent fractionation of nuclear contents by differential centrifugation was performed as described (Hügle et al., 1985; Kneissel et al., 2001). Fractions obtained were termed low-speed pellet (LSP), high-speed pellet (HSP), and high-speed supernatant (HSS). LSP fractions were cleared from yolk by extraction with freon (Evans and Kay, 1991). Unfertilized eggs from X. laevis were extracted as described (Cordes et al., 1993).
Isolation of Amplified Nucleoli by Fluorescence-Activated Particle Sorting
Mass isolated X. laevis nuclei densely packed in 5:1 medium (83 mM KCl, 17 mM NaCl, 2 mM MgCl2, 1 mM dithiothreitol, 100 µM Pefabloc, 10 mM Tris-Cl, pH 7.4) containing 100 mM sucrose were gently homogenized by sucking through a 200-µl pipette tip and labeled with fluorescein isothiocyanate (FITC)-conjugated mAb No-185 (see below, final concentration 50 µg/ml homogenate). Subsequently, the homogenate was analyzed with a fluorescence-activated cell sorting (FACS) flow cytometry system (FACS II; Becton Dickinson, Heidelberg, Germany) using the 488-nm line of an argon ion laser and 5:1 medium as sheath fluid. The relative FITC fluorescence of passing particles was measured with a 520-620-nm bandpass filter. Information about the relative particle size was obtained with a light scatter sensor. Large, strongly fluorescing particles (i.e., nucleoli) were sorted by appropriate threshold settings in the two-dimensional distribution of FITC fluorescence and light scatter (cf. Figure 1A). The collected nucleoli were sedimented (10 min, 3200 x g) and resuspended in PBS or SDS-PAGE sample buffer.
|
Protein NO66-specific Antibodies
Monoclonal antibodies against proteins from purified X. laevis oocyte nucleoli were generated essentially as described by Köhler and Milstein (1975). One of the resulting antibodies, mAb No-5-1-1 (IgG1) showed a strong nucleolar immunostaining on Xenopus A6 cells and reacted specifically with a 66-kDa polypeptide (NO66; cf. Figure 2, A and B).
|
Guinea pig antisera against protein NO66 were obtained by immunization with synthetic peptides (PSL, Heidelberg, Germany). Antisera NO66-1 and NO66-3 were raised against peptides NSSEFLPQFSSVNFR and HNHGYYDGLFSTK, respectively, representing parts of X. laevis NO66 as determined by peptide sequencing. NO66-2 was directed against the peptide TPPGTQGFAPHYDDIEAFVLQLEGKKHWRV derived from an expressed sequence tag (EST) from chicken (acc. no. AL587154
[GenBank]
). All antisera showed the same specificity and were used after affinity purification on corresponding peptide-conjugated matrices (Mertens et al., 1996).
Other Antibodies
The following antibodies were used for colocalization studies: Rabbit antibodies to coilin (Bohmann et al., 1995), PSP1 (Fox et al., 2002), fibrillarin (Azum-Gélade et al., 1994), CENP-B (Goldberg et al., 1996), XOrc1, XOrc2, XMcm3, XMcm7, and XCdc6 (Madine et al., 2000). Rabbit antiserum against human SF3b155 was obtained by immunization with baculovirus expressed recombinant protein. mAbs P7-1A4 against nucleolin and SH54 against PTB (Meßmer and Dreyer, 1993; Huang et al., 1997). mAb MB67 recognizing the Ki-67 protein and mAb PC10 against the proliferating cell nuclear antigen (PCNA) was from Dianova (Hamburg, Germany) and Dako (Hamburg, Germany), respectively. SMN, an mAb directed against the homonymous SMN protein was purchased from Becton Dickinson, mAb PML (PG-M3) against the PML protein and mAb Sam68 (7-1) against protein Sam68 were from Santa Cruz (Heidelberg, Germany), and mAb 2HP 1H5 against HP1 was from Euromedex (Mundolsheim, France). Lamin-mAb L046F7 against Xenopus oocyte-specific lamin LIII, mAb 9E10 recognizing the human c-myc protein and the guinea pig serum B2.4-1 against SF3b155 have been described (Benavente et al., 1985; Evan et al., 1985; Schmidt-Zachmann et al., 1998). mAb No-440, which recognized the nucleolar RNA helicase NOH61 (Zirwes et al., 2000), was raised against recombinant His6-tagged NOH61. mAb No-114 against xNopp180 and mAb No-185 against protein NO38/B23 were described (Schmidt-Zachmann et al., 1984, 1987).
For fluorescence-activated particle sorting, mAb No-185 was purified from serum free cell culture supernatant by affinity chromatography using HiTrap protein G-Sepharose columns (Amersham, Freiburg i. Br., Germany). The purified antibodies were covalently coupled to FITC as described by Harlow and Lane (1998).
Secondary antibodies used for immunofluorescence microscopy were Cy3- or Alexa 488-conjugated antibodies against immunoglobulin G (IgG) of mouse, guinea pig, or rabbit (Dianova). Western blot experiments were performed with horseradish peroxidase-conjugated secondary antibodies (Dianova). For immunoelectron microscopy, anti-guinea pig- or anti-mouse IgG-conjugated Nanogold (Nanoprobes, Stony Brook, NY) was used.
cDNA Clones
IMAGE Consortium (LLNL) cDNA clones (Lennon et al., 1996) with CloneIDs 4340013 (GenBank acc. no. BC011350
[GenBank]
) and 3916148 (acc. no. BC014928
[GenBank]
) were obtained from the German Resource Center for Genome Research (RZPD, Berlin). These clones contained open reading frames encoding full-length hsNO66 and hsNO52, respectively. The hsNO66-cDNA was subcloned into expression vector pcDNA3.1(+) by EcoRI/NotI, resulting in clone pcDNA3.1-NO66. pcDNA3.1-mycNO66 was constructed by digestion of pcDNA3.1-NO66 with EcoRV/NotI. A 2477-base pair cDNA fragment encoding hsNO66 was isolated and cloned in-frame with the myc-tag into the SmaI/NotIdigested construct BTmyc (Schmidt-Zachmann and Nigg, 1993). The resulting fusion construct was transferred by HindIII/NotI into pcDNA3.1(+).
RNA Isolation, Northern Blot Hybridization, and Coupled In Vitro Transcription-Translation
Total RNA from MCF-7 cells was prepared as described (Chirgwin et al., 1979). Poly(A)+ RNA was obtained using a mRNA purification kit (Amersham). RNAs (5 µg per lane) were separated on 1% agarose gels containing 0.6% formaldehyde, transferred to Biodyne B nylon filters (Pall, Dreieich, Germany) and hybridized as described (Kneissel et al., 2001) with a random-primed, 32P-labeled 554-base pair FspI/XcmI fragment of the hsNO66 cDNA.
In vitro synthesis of 35S-methionine-labeled proteins by coupled transcription/translation was carried out with the TNT reticulocyte lysate system (Promega, Mannheim, Germany).
Preparation of Total Cellular Lysates, Cell Extracts, and Isolation of Nuclei and Nucleoli from MCF-7 Cells
Total cellular lysates were prepared as described by Schmidt-Zachmann et al. (1998). Extracts from human A-431 and X. laevis A6 cells were obtained according to Zirwes et al. (2000) with lysis buffer containing 50 mM Tris-Cl, pH 7.4, 250 mM NaCl, 0.5% Triton X-100, 1 mM 
-mercaptoethanol, and 1 mM Pefabloc. The preparation of nuclei and nucleoli from MCF-7 cells was performed essentially as described by Zirwes et al. (2000).
Immunoprecipitation Experiments, Mass Spectrometry, and Peptide Sequencing
Immunoprecipitation from X. laevis A6 cell extract with mAbs was done according to Zirwes et al. (2000). Excision of immunoprecipitated protein bands from gels, digestion with trypsin, matrix-assisted laser desorption/ionization (MALDI) mass spectrometry, amino acid sequence analysis, and database searches were performed as described (Kuhn et al., 2001).
Sucrose Density Gradient Centrifugation and Gel Filtration
A-431 cell extracts and HSP fractions were separated as described (Zirwes et al., 2000) in density gradients containing 5-30% and 10-40% sucrose, respectively. In parallel experiments protein samples were preincubated with 50 µg of DNase-free RNase A for 30 min on ice before loading onto gradients.
For gel filtration, 200 µl of A-431 cellular extract corresponding to 107 cells were loaded onto a Superdex 200 HR10/30 column (Amersham) and eluted with lysis buffer (without detergent). For calibration, mixtures of dextran blue and reference proteins (thyroglobulin, ferritin, catalase, aldolase, BSA, and ovalbumin; all from Amersham) were separated under the same conditions.
Gel Electrophoresis and Immunoblotting
Protein fractions were separated by SDS-PAGE according to Thomas and Kornberg (1975), and immunoblot analysis was performed as described by Zirwes et al. (2000).
DNA Transfection and Microinjection into Cultured Cells
Transient transfection of MCF-7 cells was achieved with Fugene 6 transfection reagent (Roche Diagnostics, Mannheim, Germany) according to the manufacturer's protocol. Immunofluorescence microscopy analysis was performed 24 h after transfection. For replication foci labeling, 25 mM Cy3-dCTP in PBS was microinjected into the nucleoplasm of MCF-7 cells as described by Görisch et al. (2003). Cells were fixed and processed for immunofluorescence microscopy 45 min after injection.
Immunofluorescence Microscopy
For immunofluorescence microscopy, cells were fixed with methanol (5 min, -20°C) and acetone (30 s, -20°C). Nuclease digestion was performed according to Spector et al. (1991). In some experiments, cells were extracted with Triton X-100 before fixation as described by Fujita et al. (2002). Incubation with antibodies and DAPI staining was performed essentially as described (Zirwes et al., 2000). Dilutions in PBS of the primary antibodies raised during this study were 1:25 for NO66-2 and 1:10 for NO66-3. Concentrated hybridoma cell culture supernatant containing mAb No-440 was diluted 1:500, mAb No-5-1-1 cell culture supernatant was used undiluted.
Electron Microscopy
For conventional thin-section studies, purified X. laevis oocyte nucleoli were processed as described (Rose et al., 1995). For immunoelectron microscopy, cells grown on coverslips were fixed in 2% formaldehyde in PBS (see Figure 5I) as specified by Brandner et al. (1998) or in cold methanol and acetone (see Figure 9, F and G) as described for immunofluorescence microscopy, followed by immunogold labeling according to Zirwes et al. (2000). Antibody dilutions were the same as for immunofluorescence microscopy. For some experiments, cells were incubated with 5 µg/ml actinomycin D (AMD) for 4 h before fixation.
|
|
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONACKNOWLEDGMENTSREFERENCES |
|---|
) and Schmidt-Zachmann et al. (1984). Here, nuclear homogenates obtained from mass-isolated oocyte nuclei were incubated with FITC-coupled mAb No-185 directed against the well-characterized nucleolar protein NO38/B23 (Schmidt-Zachmann et al., 1987; Zirwes et al., 1997a). This treatment led to a strong and specific labeling of the amplified nucleoli and thus allowed their isolation by FACS (Figure 1A). The purity and integrity of the resulting fraction was analyzed by light and electron microscopy (Figure 1, B-D) as well as by immunoblot experiments (Figure 1E).
The purified amplified nucleoli, varying in size from 1 to 20 µm were highly enriched, exhibiting the typical intranucleolar morphology (Figure 1D; see also Mais and Scheer, 2001). To verify the purity of the isolated nucleoli, protein fractions from total oocyte nuclei in comparison to purified nucleoli were analyzed by SDS-PAGE, transferred to nitrocellulose filters, and probed with antibodies specific for other nucleolar proteins such as NO38, xNopp180, or nuclear proteins SF3b155 (a component of the splicing factor SF3b that associates with the U2 snRNP) and the oocytespecific lamin LIII (Figure 1E). Both nucleolar proteins xNopp180 and NO38 (Schmidt-Zachmann et al., 1984, 1987) were highly enriched in the nucleolar protein fraction, whereas the relative amounts of two other nuclear proteins, SF3b155 known to be highly enriched in nuclear speckles (Schmidt-Zachmann et al., 1998) as well as the lamina protein lamin LIII (Krohne et al., 1984) were greatly reduced in the nucleolar fraction.
The protein fraction containing amplified nucleoli in FACS-purified form was separated by two-dimensional gel electrophoresis and analyzed by mass spectrometry (MS) and microsequencing (our unpublished results). A large-scale MS-based protein identification was hampered by the lack of comprehensive Xenopus-specific protein and cDNA databases. Because the identification of the individual protein spots progressed more slowly than expected, a second approach was taken in parallel. The same protein fraction was used for the immunization of mice to generate mAbs directed against nucleolar components. Besides a number of mAbs recognizing well-known nucleolar proteins, such as NO38, xNopp180, and nucleolin, a mAb termed No-5-1-1 was obtained that showed a strong and specific labeling of nucleoli in somatic cells (Figure 2A) as well as in oocytes from X. laevis (our unpublished results). Immunoprecipitations from protein extracts prepared from XLKE-A6 cells using mAb No-5-1-1 revealed the enrichment of a ca. 66-kDa polypeptide (Figure 2B, lane 3). To identify this polypeptide, hereafter referred as to NO66, the immunoprecipitated band was excised from the gel and digested with trypsin. The resulting peptides were analyzed by MALDI-TOF mass spectrometry, and the peptide mass fingerprints were compared with predicted peptides of the NCBInr database by the ProFound search aligorithm. As the mass information obtained for the 66-kDa protein could not be assigned to any known protein in the database, we prepared material sufficient for microsequencing analysis of the polypeptide.
Twelve amino acid sequences were obtained (Figure 2C) and subsequently used for detailed database searches. We identified a human cDNA clone encoding the hypothetical protein FLJ21802displaying significant homology to various peptides derived from the Xenopus protein NO66 (thereafter termed hsNO66; Figure 3A). The human cDNA clone was received from the RZPD and used for further studies of this novel nucleolar protein (acc. no. AY390535
[GenBank]
). The cDNA clone coding for hsNO66 (2451 base pairs) contained an initiation codon at position 52, an open reading frame of 1926 base pairs and a 3'-untranslated region of 474 base pairs. The open reading frame encoded a polypeptide of 641 aa with a calculated molecular mass of 71.1 kDa and an isoelectric point of 6.4. Inspection of the primary sequence of hsNO66 disclosed the presence of two putative nuclear localization signals, one of the monopartite type (aa 17-21; RRRRK) and the other of the bipartite type (aa186-201; RR (x)11 RRR; cf. Dingwall and Laskey, 1991). A high number of potential phosphorylation sites for cAMP-dependent kinase, protein kinase C, casein kinase II, and thyrosine kinase were also noticed. Presently, we do not know the actual degree of phosphorylation of NO66, but posttranslational modifications could well account for the observed difference in molecular mass calculated for the polypeptide from the cDNA-derived sequence (71 kDa) or from its SDS-PAGE mobility (79 kDa, see below). A closer analysis of the primary sequence of the protein revealed the presence of a putative JmjC domain (aa 290-429). This kind of domain has been found in numerous eukaryotic proteins and has been postulated to be involved in chromatin remodeling processes (Clissold and Ponting, 2001).
|
Searches of current databases have disclosed a striking homology of hsNO66 with hypothetical proteins from Anopheles gambiae, C. elegans, Drosophila melanogaster, Rattus norvegicus, and Mus musculus (acc. nos. EAA01582 [GenBank] T15138 [GenBank] , NP_572160 [GenBank] , Xp_234396, and BAA95038 [GenBank] respectively) as well as with expressed sequence tags of Gallus gallus (Figure 3B; acc. no. AL587154 [GenBank] ), and Xenopus tropicalis (Figure 3C; acc. nos. AL650484 [GenBank] and AL786378 [GenBank] ). These proteins of unknown functions comprise of 693, 817, 653, 597, and 603 aa residues, and differ markedly in size from the human protein, because of different lengths of the amino-terminal portions ranging from 147 aa (M. musculus) to 361 aa (C. elegans). Notably, all proteins show a significantly high sequence homology upstream of the less conserved amino-terminal third, corresponding to aa 186-641 of hsNO66. Here, the sequence identity between hsNO66 and the homologues in A. gambiae, C. elegans, D. melanogaster, and M. musculus are 44, 39, 40, and 92%, respectively (our unpublished results). This remarkable conservation indicated that these proteins were indeed homologous and most likely participate in a fundamental cellular process.
Moreover, database searches revealed the existence of another related but distinct hypothetical human protein (FLJ14393 with a molecular mass of 52 kDa (termed NO52; acc. no. AY390536
[GenBank]
), which might be correlated to the recently described Mina53 protein proposed to be involved in cell proliferation (Tsuneoka et al., 2002). The overall aa sequence identity between the two human proteins was 36% but even higher in a central domain of 140-aa residues (57%). Like protein NO66, protein NO52 was also highly conserved during evolution and was found enriched in nucleoli. Further studies, however, revealed that the two nucleolar proteins differed in their biochemical properties as well as in the ways their precise intracellular localization is affected by certain treatments such as serum starvation and transcription inhibition by AMD (our unpublished results; a detailed characterization of protein NO52 will be presented elsewhere).
Because for unknown reasons all attempts to clone the X. laevis ortholog failed, we used the cDNA clone coding for the human protein (hsNO66) for the molecular and biochemical characterization of this novel nucleolar protein.
Molecular Characterization of Protein NO66
Total RNA as well as poly(A+) RNA from human MCF-7 cells were probed in Northern blot experiments, using a random prime-labeled fragment of the hsNO66-cDNA. Although only a faint signal could be observed in total cellular RNA after long exposure, a specific signal corresponding to a mRNA of 2.3 kb was detected in the poly(A+) fraction, indicating that the analyzed cDNA clone encoding protein NO66 was of full length and that the amount of the endogenous mRNA encoding this nucleolar protein was rather low in the cells used (Figure 3D). When the expression of the gene encoding protein NO66 was examined, the 2.3-kb mRNA was detected in all eight tissues (heart, brain, placenta, lung, liver, skeletal muscle, kidney, and pancreas), although the intensity of the radioactive signal varied somewhat (our unpublished results). To analyze the localization of hsNO66, a chimeric cDNA construct encoding the full-length protein carrying a myc tag at its amino-terminus was transfected into MCF-7 cells. Immunofluorescence analysis with the myc antibody revealed that the protein was highly enriched in nucleoli. However, additional labeling of small nucleoplasmic granules was also observed (Figure 3, E and E').
Because mAb No-5-1-1 recognizing X. laevis NO66 did not cross-react with other species, except for X. borealis, as indicated from the immunofluorescence studies, and was negative in immunoblots, we generated further antibodies to study the intracellular distribution and localization of the endogenous protein hsNO66. Three polyclonal antibodies (NO66-1, -2 and -3) raised against peptides representing different sequence elements of protein NO66 were obtained, which varied in their cross-species reactivities but showed the same specificity. To prove that the antisera indeed recognized the same protein originally detected by mAb No-5-1-1 in cellular extracts from XLKE-A6 cells, we repeated the immunoprecipitation shown in Figure 2B and analyzed the resulting protein fractions by immunoblotting using these polyclonal NO66 antibodies (Figure 4). The antibodies showed a strong reaction with the Xenopus protein present in the total cellular extract before and after incubation with protein G-Sepharose in the absence of antibody No-5-1-1 (Figure 4B, lanes 1 and 2), which was no longer detectable in the extract upon incubation with protein G-Sepharose in the presence of the antibody. Moreover, the NO66 antibodies specifically reacted with the polypeptide immunoprecipitated in Coomassie blue-visible amounts by the mAb (Figure 4B, lane 5). From these results, we concluded that the hsNO66 cDNA clone coded for the human homologue of the Xenopus protein NO66, originally identified with the mAb No-5-1-1. We also verified that the NO66 antisera did not cross-react with the sequence-related 52 kDa polypeptide mentioned before (our unpublished results).
|
In immunoblots of total proteins isolated from different cultured human cells antibody NO66-2 reacted exclusively with a polypeptide of 79 kDa (Figure 5, A and A'), in good agreement with the size of the product of in vitro translation of the hsNO66 cDNA (our unpublished results). Although the antigenic polypeptide was present in similar amounts in most cell lines analyzed, it was clearly underrepresented in the cells of the glioma line U333 (Figure 5A', lane 2). Similar analyses of protein fractions derived from diverse species (mouse, rat, bovine, canine, rat kangaroo, chicken, trout, and Xenopus) disclosed the presence of the antigen with molecular weights ranging from 55 kDa (trout) to 80 kDa (potoroo), reflecting variations of the length of the amino-termini (Figure 5, B and B'). The broad cross-reaction observed again emphasized that protein NO66 has been highly conserved during evolution.
Intracellular Distribution of Protein NO66
We next examined the intracellular localization of native protein NO66 by immunofluorescence microscopy on the same monolayer cell cultures, which were all analyzed at approximately the same cell density (80% confluence). Surprisingly, significant differences in the intracellular staining pattern could be observed that conformed to one of five patterns of distribution as summarized in Table 1. In most cell lines studied, NO66 antibodies brightly stained the nucleoli, the nucleoplasm in a finely dispersed form and, additionally, some distinct nuclear "dots," some of which were often seen in association with nucleoli (pattern I in Table 1). A weak nucleolar fluorescence, but prominent nucleoplasmic dots were observed in other cells such as those of lines PLC and BMGE. Poorly stained nucleoli and diffuse labeling of the nuclear interior were seen in other cell types (RV, PtK2, and RTG-2). The next class of intracellular distribution pattern was defined by a rather diffuse nucleoplasmic staining that, however, excluded the nucleoli (3T3), and the glioma line U333 showed strongly labeled nuclear dots, but again no nucleolar fluorescence (Figure 5, C and C'-H and H'; and our unpublished results).
|
Because the majority of cells showed a strong nucleolar accumulation of protein NO66, we investigated the intranucleolar distribution of the protein in greater detail by performing immunoelectron microscopy on AMD-treated XLKE-A6 cells. Under these conditions the nucleolar subcompartments became separated from each other, thereby facilitating intranucleolar localization. As shown in Figure 5I, protein NO66 was significantly enriched in the granular component, whereas the dense fibrillar component was practically devoid of gold particles. The same intranucleolar distribution was observed in untreated cells (our unpublished results), indicating that protein NO66 may have a role in later stages of the processing of preribosomal particles, at least in some cell types.
Many nuclear complexes disperse in mitotic cells, but some remain in association with the segregating chromosomes. Interestingly, the intracellular distribution of protein NO66 changes drastically during mitosis. During meta- and early anaphase the protein is diffusely distributed over the perichromosomal cytoplasm, whereas during late anaphase a significant fluorescence is seen on the surfaces of the chromosomes. Later in telophase, the protein appears to be enriched in prenucleolar bodies (our unpublished results). This pattern of distribution during mitosis is similar, but not identical to the mitotic behavior described for other nucleolar proteins such as NO38, NO29, and NOH61, which associate with the chromosomes already during metaphase (Schmidt-Zachmann et al., 1987; Zirwes et al., 1997b, 2000).
Notably, the immunolocalization pattern of protein NO66 was unaffected after incubation the cells with DNase but was drastically altered by treatment with RNase, resulting in a complete release from the nucleoli as well as from the nucleoplasmic structures (our unpublished results). On treatment of cells with AMD protein NO66 was to some extent released from the nucleoli and translocated to the nucleoplasm. These experiments indicated that protein NO66 binds to RNA and that its nucleolar accumulation depends on ongoing transcription.
Biochemical Characterization and Immunoprecipitation Experiments
Our transfection experiments (Figure 3, E and E') and immunolocalization studies (Figure 5) had demonstrated a mostly nucleolar localization of protein NO66. To examine this on the biochemical level, we analyzed the distribution of protein NO66 in cellular fractions from cultured human cells by gel filtration and sucrose gradient centrifugation.
On cell extractions with different detergent and salt concentrations, the protein was largely solubilized in buffers containing 250 mM NaCl and 0.1% Triton X-100 (our unpublished results). On gel filtration the solubilized protein appeared with an apparent molecular mass (Mapp) of 600,000 (Figure 6, A and A') and was identified in a complex with a sedimentation coefficient of 12S (Figure 6, B and B') when analyzed by sucrose gradient centrifugation. According to Siegel and Monty (1966), this allowed to estimate a molecular mass of 350,000 Da for the solubilized form of protein NO66. RNase treatment of the cellular lysates before centrifugation led to a significant degradation of the protein as indicated by the much weaker signal obtained by immunoblotting. The remaining protein seemed to form large aggregates, which sedimented at the bottom of the gradient (Figure 6C). These results suggested that native protein NO66 is a constituent of a larger particle containing an RNA component of so far unknown identity.
|
On biochemical fractionation of MCF-7 cells protein NO66 was absent from the cytoplasmic fraction but significantly enriched in the nuclear and nucleolar fraction (Figure 6, D and D'). The fractionation procedure was verified by immunoblot analyses with antibodies B2.4-1 directed against the nuclear protein SF3b155 (Figure 6D'') as well as with antitubulin, anti-NOH61, and antilamin B2 (our unpublished results).
The distribution of protein NO66 in different fractions from Xenopus oocyte nuclei, egg extracts, and somatic cells was analyzed by immunoblotting with NO66-specific antibodies. Here, the protein was detected in total nuclei as well as in the LSP and HSP fraction, indicating that NO66 was a nuclear protein bound to relatively large structures. By contrast, the protein was not detected in the HSS fraction and present only in trace amounts in egg extracts (Figure 6, E and E'). Protein NO66 was also highly enriched in the fraction containing amplified nucleoli obtained by particle sorting (our unpublished results). As the NO66-containing HSP fraction was enriched in preribosomes (Hügle et al., 1985), we fractionated it further on sucrose gradients and analyzed the resulting fractions by SDS-PAGE and immunoblotting (Figure 6F). Protein NO66 was specifically enriched in fractions containing the precursors for the large ribosomal subunit and even larger particles, probably representing intermediates of the 90 S preribosomal particle (Figure 6F'). By contrast, the major nucleolar protein NO38 was specifically enriched in 40S and 65S particles representing precursors for the small and large ribosomal subunits (Figure 6F''; cf. Schmidt-Zachmann et al., 1987). Again, RNase treatment led to an almost complete degradation of protein NO66 with the remaining protein sedimenting at the bottom of the gradient (Figure 6G), whereas protein NO38 was released from the ribonucleoprotein particles and shifted to the top of the gradient (Figure 6G').
In conclusion, the biochemical characterization of NO66 indicates that this protein represents a factor associated with preribosomal complexes and may accompany preribosomes through several stages of maturation.
A number of nucleolar proteins have been implicated directly or indirectly in pre-rRNA packaging and metabolism in higher eukaryotic cells such as protein NO38/B23, the box C/D snoRNA-associated protein fibrillarin, protein Nopp140 (xNopp180 in Xenopus) transiently interacting with both major classes of snoRNPs, box H/ACA and box C/D, the nucleolar helicase NOH61, and the very acidic protein NO29 (reviewed by Olson et al., 2002). Therefore, we determined whether these proteins are among those that copurify with protein NO66 by immunoblot analysis (Figure 7). Indeed, nucleolin (Figure 7A') and NO38/B23 (Figure 7A'') were detected in the complexes isolated with mAb No-5-1-1. An interaction between protein NO66 and nucleolin as well as protein NO38 was confirmed in reciprocal immunoprecipitations using nucleolin- and NO38-specific antibodies, respectively (Figure 7, B and C). In both cases, protein NO66 was found exclusively in the fractions containing the purified immunocomplexes (Figure 7, B' and C'). In contrast, none of the other aforementioned nucleolar proteins (xNopp180, NO29, NOH61, fibrillarin) was detectable in complexes with protein NO66 (our unpublished results). Therefore, we concluded that only a subset of nucleolar proteins are found in association with protein NO66, which is in agreement with recent reports of a large number of preribosomal particles of different molecular composition (reviewed by Tschochner and Hurt, 2003).
|
Characterization of the Protein NO66-Containing Complexes by Colocalization with Nuclear Marker Proteins
As mentioned earlier, protein NO66 displayed cell type-dependent differences of intranuclear localization (Table 1), raising the question whether its location is unique or overlaps with other nuclear entities (for recent reviews of distinct nuclear domains see Lamond and Earnshaw, 1998; Misteli, 2000). By confocal laser scanning microscopy, we directly compared the localization of protein NO66 with that of well-characterized marker proteins of different kinds of "nuclear bodies," in particular the locations of nucleoli (Figure 8, A-A'''), Cajal bodies (Figure 8, B-B'''), gems (Figure 8, C-C'''), PML bodies (Figure 8, D-D'''), nuclear speckles (Figure 8, E-E'''), paraspeckles (Figure 8, F-F'''), or the perinucleolar compartment (PNC; Figure 8, G-G'''). Clearly, protein NO66 did not colocalizes with NOH61 in dots outside nucleoli. In addition, subnuclear structures containing coilin, SMN, PML, SF3b155, PSP1, and PTB protein all showed no significant codistribution with protein NO66. Moreover, the NO66-positive nuclear dots were clearly different from Sam68 domains as well as from transcription sites (our unpublished results). In summary, these results demonstrated that protein NO66 can occur besides nucleoli in unique nucleoplasmic locations distinct from the major subnuclear domains involved in RNA processing and transcriptional regulation.
|
Because of similarities in the intranuclear distribution between NO66-positive nuclear dots and another type of nuclear bodies, the replication foci, we microinjected Cy3-dCTP into the nuclei of MCF-7 cells. After costaining with NO66-antibodies we observed a significant overlap between the protein NO66-containing structures and the sites of Cy3-dCTP incorporation (Figure 9, A-A'''). This positive correlation was further confirmed by colocalizing protein NO66 with the proliferating cell nuclear antigen (PCNA; Figure 9, B-B'''), a processivity factor of DNA polymerase 
(reviewed by Wyman and Botchan, 1995; Hingorani and O'Donell, 2000). In contrast to PCNA, which was removed from non-S-phase cells treated with Triton X-100 (Sasaki et al., 1994; Lohr et al., 1995), protein NO66 remained in non-S-phase cells and showed the same intranuclear distribution as in S-phase cells (Figure 9, C-C'''). The nuclear sites that had incorporated Cy3-dCTP and were positive for PCNA were rather large and often were located in the nuclear periphery or in association with nucleoli. This pattern closely resembled replication foci described for cells in mid-to-late S-phase, i.e., when switching from the replication of euchromatin to that of heterochromatin (Leonhardt et al., 2000). Because PCNA-containing DNA replication foci undergo characteristic changes as replication progresses, we examined the temporal and spatial relationship between NO66 and PCNA foci during S-phase progression in synchronized cells. No colocalization could be observed at the G1/S border (time point 0 h) or in early to mid-S-phase (time points 2-6 h after release). In contrast, 8 h after the cells had been released, i.e., in late S-phase, the two proteins were substantially colocalized (Figure 9, D-D'''; for the complete analysis see Supplemental Figures, Figure 1s). Notably, the intracellular concentration of protein NO66 was not altered during the cell cycle stages analyzed, which argued against a direct functional role of the protein in cell cycle regulatory processes (Supplemental Figures, Figure 2s). Thus, colocalization of NO66 and replication foci were demonstrated exclusively in late S-phase cells, suggesting that this protein marks distinct heterochromatic regions.
Another nucleolar protein known to also occur in nuclear dots representing specific heterochromatin domains is the proliferation marker protein Ki-67 (Bridger et al., 1998). In double localization studies using the appropriate antibodies, we have indeed observed a striking, although not complete colocalization of these two proteins (Figure 10, A-A'''). Well-known marker proteins for distinct heterochromatic regions are the proteins CENP-B and HP1. whereas CENP-B is distributed throughout the centromeric heterochromatin by binding to a distinct DNA sequence located in the repeats of -satellite DNA (Cooke et al., 1990), the HP1 protein is a constituent of pericentromeric heterochromatin (reviewed by Eissenberg and Elgin, 2000). Comparison of the intranuclear distribution of NO66 with that of CENP-B has not disclosed a significant overlap of the respective foci, although NO66- and CENP-B-positive subnuclear structures are often found in close proximity (Figure 10, B-B'''). In contrast, when NO66 has been compared with HP1, striking colocalization has been observed in all cell lines examined, which, however, has been incomplete because the number of HP1 foci is higher than the number of NO66 foci. Remarkably, almost every NO66-positive nucleoplasmic structure also presents local accumulation of HP1 (Figure 10, C-E).
|
In summary, protein NO66 displays a characteristic dual location. Besides its nucleolar accumulation, it is also enriched in nuclear foci that most probably correspond to a certain type of heterochromatin, which is replicated during late stages of S-phase.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONACKNOWLEDGMENTSREFERENCES |
|---|
66 kDa) with a hitherto unknown localization. Protein NO66 is present in various cell types of diverse species from amphibian to human, where it appears with different molecular weights due to variations in length of the amino-terminal region. Both, the high evolutionary conservation in aa sequence between the human protein and putative counterparts in C. elegans and X. tropicalis as well as the wide cross-reactivity of NO66 antibodies suggest that this protein serves a fundamental cellular process. Remarkably, however, no homologue is found in yeast, and the protein is not included in the recently published lists of nucleolar proteins identified in the course of proteomic analyses of human nucleoli (Andersen et al., 2002; Scherl et al., 2002).
The primary sequence of protein NO66 did not disclose obvious functional roles, and no well-characterized motifs were detectable. However, NO66 contains a region showing striking similarity to the COG2850 (Cluster of Orthologous Groups of proteins; http://www.ncbi.nlm.nih.gov/BLAST) consensus sequence that defines a group of 49 otherwise uncharacterized conserved proteins. In addition to NO66 this group comprises several predictable bacterial proteins, hence underscoring the high evolutionary conservation of protein NO66. Interestingly, all these proteins also exhibit a JmjC-domain and are candidates for enzymes involved in chromatin remodeling, thereby regulating gene expression (Clissold and Ponting, 2001; Ayoub et al., 2003).
The intranuclear topology of protein NO66 is as characteristic as perplexing and can differ in different kinds of cells. Typically, it occurs in more than one structural entity and thus can be counted among the "dual location-dual functional candidate" proteins (for review see Smalheiser, 1996). In light and electron microscopic immunolocalization studies five main patterns of intranuclear distribution have been distinguished, depending on the specific cell type (Table 1). However, in the majority of cell types studied, protein NO66 is strongly enriched in nucleoli but additionally occurs in distinct nucleoplasmic foci, often located in close proximity to the nucleoli. Such different distribution patterns may reflect either regulations by intracellular signals, including conformational changes induced by posttranslational modifications or differences in the expression level. Surprisingly, transient transfection experiments performed in various cell lines (MCF-7, PLC, 3T3) disclose the almost exclusive nucleolar localization of ectopically expressed protein NO66 without affecting the nucleoplasmic location of the endogenous protein. Moreover, photobleaching experiments have demonstrated that many nuclear proteins move constantly in and out of a given nuclear substructure such as the nucleolus (Chen and Huang, 2001), with transient accumulations in other structural entities (Leung and Lamond, 2003). Thus, protein NO66 might shuttle between the nucleolus and the distinct nuclear dots, and this equilibrium may be differently regulated in the various cell types.
Our biochemical data also justify to consider protein NO66 as a constitutive nucleolar component: The protein is detectable in purified extrachromosomal nucleoli of Xenopus oocytes as well as in chromosomal nucleoli of somatic cells. In addition, our finding that it is associated with large preribosomal particles (>60S) and coimmunoprecipitates with two nucleolar hallmark proteins, NO38 and nucleolin, suggests that NO66 plays a role in the assembly and processing of ribosomal subunits, notably of early 65S-90S preribosomal complexes. Such a role would also be in agreement with its localization to the granular component as the site of maturation of preribosomal particles and would also be compatible with the demonstrated RNase effects shown. Here, the unexpected observation that protein NO66, in contrast to other nucleolar proteins known to be associated with preribosomes, does not dissociate under the influence of RNase action but rather forms aggregates reminds one of earlier studies reviewed by Pederson (2000), showing that nuclear RNA-bound proteins can rearrange to large particles when dislodged from their usual RNA associations.
In most cell culture lines examined, protein NO66 did not exclusively localize to nucleoli but also accumulated in certain extranucleolar, nucleoplasmic bodies. To identify the nature of these intranuclear structures, a number of colocalization studies have been performed using antibodies against appropriate marker proteins for well-established nuclear entities such as Cajal bodies, speckles, PML bodies, paraspeckles, and the perinucleolar compartment. However, none of these proteins shows colocalization with protein NO66. On the other hand, the nucleoplasmic staining pattern of NO66, although persistent over the entire interphase, is somewhat reminiscent of DNA replication foci. It is well known that the pattern of DNA replication sites within the nucleus changes in many cultured cells from a finely punctate distribution in early S-phase cells, followed by the accumulations of larger granules at the nucleolar periphery and the nuclear envelope in mid-S-phase, to finally, the appearance of large foci in late S-phase (Leonhardt et al., 2000). Moreover, no fundamental differences in the distribution of DNA replication sites and the spatiotemporal sequence of replication patterns have been found in normal, immortalized and transformed cells (Dimitrova and Berezny, 2002). Indeed, we have detected a significant overlap between the NO66-positive dots and the replication sites labeled by incorporated C
