Nucleolar localization of parafibromin is mediated by three nucleolar localization signals

Parafibromin is a putative tumor suppressor encoded by HRPT2 and implicated in parathyroid tumorigenesis. We previously reported a functional bipartite nuclear localization signal (NLS) at residues 125-139. We now demonstrate that parafibromin exhibits nucleolar localization, mediated by three nucleolar localization signals (NoLS) at resides 76-92, 192-194 and 393-409. These NoLS represent clusters of basic amino acids arginine and lysine, similar to those found in other nucleolar proteins, as well as being characteristic of NLSs. While parafibromin's bipartite NLS is the primary determinant of nuclear localization, it does not mediate nucleolar localization. In contrast, the three identified NoLSs play only a minor role in nuclear localization, but are critical for the nucleolar localization of parafibromin.     

Conformational determinants of the intracellular localization of midkine

Midkine (MK) is a multifunctional growth factor and has been discovered to play important roles in carcinogenesis. MK has been reported to localize to the nucleus and nucleolus, however, the data are not consistent and the signals responsible for the localization are unknown. Here we reported that human MK exclusively localized to the nucleus and nucleolus in HepG2 cells by using GFP as a tracking molecule. In order to identify the motifs required for the nuclear localization and nucleolar accumulation, point- and deletion-mutations were introduced and the corresponding subcellular localizations were analyzed. Data revealed that (i) K79R81, K86K87, and the C-terminal tail of MK constitute the nuclear localization determinant of MK, and (ii) the C-terminal tail is the key element controlling MK nucleolar accumulation though the N-terminal tail, K79R81, and K86K87 also contribute to this process. Taken together, our results provide the first documentation about the determinants required for MK nuclear and nucleolar localization.     

KIAA0649, a 1A6/DRIM-interacting protein with the oncogenic potential

We identified a 1A6/DRIM (down-regulated in metastasis) interacting protein, KIAA0649 during the yeast two-hybrid screen. The interaction between KIAA0649 and 1A6/DRIM was further confirmed by GST-pull-down and co-immunoprecipitation assays. KIAA0649 was originally identified from human brain tissue. However, its biological function remains unknown. In this report, we showed that KIAA0649 mRNA is widely expressed in human multiple tissues and cell lines. We have also demonstrated that KIAA0649 has oncogenic characteristics: it enhances colony formation, allows anchorage-independent growth. Moreover, KIAA0649 exogenous expression in NIH3T3 fibroblasts caused tumor development in nude mice.     

Nuclear and nucleolar localization signals and their targeting function in phosphatidylinositol 4-kinase PI4K230

PI4K230, an isoform of phosphatidylinositol 4-kinase, known primarily as a cytoplasmic membrane-bound enzyme, was detected recently also in the nucleolus of several cells. Here we provide mechanistic insight on the targeting function of its putative nuclear localization signal (NLS) sequences using molecular modeling, digitonin-permeabilized HeLa cells and binding to various importins. The synthetic sequence ⁹¹⁶NFNHIHKRIRRVADKYLSG⁹³⁴ comprising a putative monopartite NLS (NLS1), targeted covalently bound fluorescent BSA to the nucleoplasm via classical importin α/β mechanism employing importins α1 and α3 but not α5. This transport was inhibited by wheat germ agglutinin and GTPγS. The sequence ¹⁴¹⁴SKKTNRGSQLHKYYMKRRTL¹⁴³³, a putative bipartite NLS (NLS2) proved ineffective in nuclear targeting if conjugated to fluorescently labeled BSA. Nonetheless, NLS2 or either of its basic clusters directed to the nucleolus soybean trypsin inhibitor that can pass the nuclear pore complex passively; moreover, an expressed 58 kDa fragment of PI4K230 (AA1166–1667) comprising NLS2 was also imported into the nucleus by import factors of reticulocyte lysate or by importin α1/β or α3/β complexes and localized to the nucleolus. We conclude that the putative bipartite NLS itself is a nucleolar targeting signal, and for nuclear import PI4K230 requires a larger sequence around it or, alternatively, the monopartite NLS.     

Identification of a nuclear localization signal in the retinitis pigmentosa-mutated RP26 protein, ceramide kinase-like protein

Retinitis pigmentosa (RP) is a genetically heterogeneous disease characterized by degeneration of the retina. A mutation in a new ceramide kinase (CERK) homologous gene, named CERK-like protein (CERKL), was found to cause autosomal recessive retinitis pigmentosa (RP26). Here, we show a point mutation of one of two putative nuclear localization signal (NLS) sequences inhibited the nuclear localization of the protein. Furthermore, the tetra-GFP-tagged NLS, which cannot passively enter the nucleus, was observed not only in the nucleus but also in the nucleolus. Our results provide the first evidence of the active nuclear import of CERKL and suggest that the identified NLS might be responsible for nucleolar retention of the protein. As recent studies have shown other RP-related proteins are localized in the nucleus or the nucleolus, our identification of NLS in CERKL suggests that CERKL likely plays important roles for retinal functions in the nucleus and the nucleolus.     

Mapping and characterization of the functional domains of the nucleolar protein RNA helicase II/Gu

RNA helicase II/Gu (RH-II/Gu) is a nucleolar RNA helicase of the DEAD-box superfamily. In this study, the functional domains of RH-II/Gu molecule were mapped by fusing the protein or its deletion mutants with a green fluorescence protein and subsequently transfecting or microinjecting the recombinant constructs into HeLa cells. In addition to the identification of a nuclear localization signal (NLS) in the N-terminus and a nucleolar targeting signal in the central helicase domain, a hidden NLS and a nucleolar targeting signal were found in the C-terminal arginine/glycine-rich domain. RH-II/Gu colocalized with fibrillarin, a component of the dense fibrillar region of the nucleolus. Overexpression of the entire RH-II/Gu protein or specific domains of the protein in HeLa cells did not interfere with the normal distribution of fibrillarin. However, when the helicase domain was truncated, the distribution pattern of fibrillarin was distorted. Microinjection of the wild-type RH-II/Gu cDNA into the nucleus of HeLa cells did not disrupt normal cell growth. However, when cells were injected with mutant DNA, only a small percentage of HeLa cells progressed through the cell cycle. Analysis of centrosomes in transfected cells demonstrated that most of the mutant-expressing cells were arrested early in the cell cycle. The results suggest that each of the structural domains of RH-II/Gu is necessary for cell growth and cell cycle progression.     

Regulation of subcellular localization of the antiproliferative protein Tob by its nuclear export signal and bipartite nuclear localization signal sequences

Tob, a member of the Tob and BTG antiproliferative protein family, plays an important role in many cellular processes including cell proliferation. In this study, we have addressed molecular mechanisms regulating subcellular localization of Tob. Treatment with leptomycin B, an inhibitor of nuclear export signal (NES) receptor, resulted in a change in subcellular distribution of Tob from its pan-cellular distribution to nuclear accumulation, indicating the existence of NES in Tob. Our results have then identified an N-terminal region (residues 2-14) of Tob as a functional NES. They have also shown that Tob has a functional, bipartite nuclear localization signal (NLS) in residues 18-40. Thus, Tob is shuttling between the nucleus and the cytoplasm by its NES and NLS. To examine a possible relationship between subcellular distribution of Tob and its function, we exogenously added a strong NLS sequence or a strong NES sequence or both to Tob. The obtained results have demonstrated that the strong NLS-added Tob has a much weaker activity to inhibit cell cycle progression from G0/G1 to S phase. These results suggest that cytoplasmic localization or nucleocytoplasmic shuttling is important for the antiproliferative function of Tob.     

Regulation of the localization and stability of Cdc6 in living yeast cells

The Cdc6 protein is an essential regulator for initiation of DNA replication. Following the G1/S transition, Cdc6 is degraded through a ubiquitin-mediated proteolysis pathway. In this study, we tagged Cdc6 with green fluorescent protein (GFP) and used site-specific mutations to study the regulation of Cdc6 localization and degradation in living yeast cells. Our major findings are: (1). Cdc6-GFP distributes predominantly in the nucleus in all cell cycle stages, with a small increase in cytoplasmic localization in G2/M cells. (2). This nuclear localization is critical for Cdc6 degradation. When the N-terminal nuclear localization signal (NLS) was mutated, Cdc6-GFP no longer accumulated in the nucleus, and the mutant cdc6 was stabilized compared to wild type. (3). The putative CDK phosphorylation sites are not required for Cdc6 nuclear localization, but are important for protein stability. These observations suggest that the stability of Cdc6 protein is regulated by two factors: nuclear localization and phosphorylation by CDK1.     

Regulation of nucleocytoplasmic localization of the atDjC6 chaperone protein

Summary. The sequence of the atDjC6 chaperone protein includes three potential nuclear localization signal (NLS) sequences (A–C) and three potential nuclear export signal (NES) sequences (X–Z). The subcellular localization of atDjC6 was studied by scanning laser confocal microscopy of chimera with the green-fluorescent protein (GFP) transiently expressed in tobacco BY-2 cells. The localization of the atDjC6::GFP chimera was coincident with that of the nuclear stain propidium iodide. Site-directed mutagenesis was used to verify the predicted NLS sequences. Each was individually fused to GFP and tested for protein localization. The individual NLS sequences were sufficient to direct partial nuclear localization of GFP, although the targeting information within NLS-B is apparently conformation sensitive. Site-directed mutagenesis of the NES sequences increased the amount of each chimera that was nuclearly localized, indicating a decrease in nuclear export. When any pair of NLS sequences were appended to GFP, the chimera were entirely nuclearly localized. Quantitative two-hybrid analysis was used to verify that the decoding of NLS sequence information involves interaction with the NLS-receptor protein importin-. Each of the NLS sequences is flanked by a site of potential Ser phosphorylation, and recombinant atDjC6 could be phosphorylated in vitro. Mutagenesis of Ser residues to the P-Ser mimic Asp interfered with nuclear targeting, apparently by preventing recognition or binding by importin-. Our results are consistent with a regulated nucleocytoplasmic localization of the atDjC6 chaperone protein.Correspondence and reprints: Plant Genetics Research Unit, USDA Agricultural Research Service, 108 Curtis Hall, University of Missouri, Columbia, MO 65211, U.S.A.     

Induction of apoptosis by NORE1A in a manner dependent on its nuclear export

The RASSF family proteins were identified as tumor suppressors in a variety of human cancers, and evidenced distinct subcellular localization patterns among their subfamilies and isoforms. In this study, we showed that NORE1A was exported actively via its nuclear export signal (NES) in the C-terminus (residues 372-379). Substitutions of three lysine residues of NORE1A NES to alanines (L372, 376, 379A) showed its localization to the dot structures of the nucleus, which was similar to the NORE1A localizations observed after the administration to cells of Leptomycin B, a nuclear export inhibitor. The NORE1A NES mutant inhibited caspase-mediated apoptosis, whereas wild-type NORE1A induced caspase-3 activation. Furthermore, the NORE1A NES mutant did not co-localize with GFP-MST1, the direct downstream target of NORE1A. These results show that the nuclear export of NORE1A via NES is involved in the NORE1A-mediated induction of apoptosis.     

Nucleolar localization/retention signal is responsible for transient accumulation of histone H2B in the nucleolus through electrostatic interactions

The majority of known nuclear proteins are highly mobile. The molecular mechanisms by which they accumulate inside stable compartments that are not separated from the nucleoplasm by membranes are obscure. The compartmental retention of some proteins is associated with their biological function; however, some protein interactions within distinct nuclear structures may be non-specific. The non-specific retention may lead to the accumulation of proteins in distinct structural domains, even if the protein does not function inside this domain. In this study, we have shown that histone H2B-EGFP initially accumulated in the nucleolus after ectopic expression, and then gradually incorporated into the chromatin to leave only a small amount of nucleolus-bound histone that was revealed by removing chromatin-bound proteins with DNase I treatment. Nucleolar histone H2B had several characteristics: (i) it preferentially bound to granular component of the nucleolus and interacted with RNA or RNA-containing nucleolar components; (ii) it freely exchanged between the nucleolus and nucleoplasm; (iii) it associated with the nuclear matrix; and (iv) it bound to interphase prenuclear bodies that formed after hypotonic treatment. The region in histone H2B that acts as a nucleolar localization/retention signal (NoRS) was identified. This signal overlapped with a nuclear localization signal (NLS), which appears to be the primary function of this region. The NoRS activity of this region was non-specific, but the molecular mechanism was probably similar to the NoRSs of other nucleolar proteins. All known NoRSs are enriched with basic amino acids, and we demonstrated that positively charged motifs (nona-arginine (R9) and nona-lysine (K9)) were sufficient for the nucleolar accumulation of EGFP. Also, the correlation between measured NoRS activity and the predicted charge was observed. Thus, NoRSs appear to achieve their function through electrostatic interactions with the negatively charged components of the nucleolus. Though these interactions are non-specific, the functionally unrelated retention of a protein can increase the probability of its interaction with specific and functionally related binding sites.     

Evolutionary conservation of nuclear and nucleolar targeting sequences in yeast ribosomal protein S6A

Over 1 billion years ago, the animal kingdom diverged from the fungi. Nevertheless, a high sequence homology of 62% exists between human ribosomal protein S6 and S6A of Saccharomyces cerevisiae. To investigate whether this similarity in primary structure is mirrored in corresponding functional protein domains, the nuclear and nucleolar targeting signals were delineated in yeast S6A and compared to the known human S6 signals. The complete sequence of S6A and cDNA fragments was fused to the 5'-end of the LacZ gene, the constructs were transiently expressed in COS cells, and the subcellular localization of the fusion proteins was detected by indirect immunofluorescence. One bipartite and two monopartite nuclear localization signals as well as two nucleolar binding domains were identified in yeast S6A, which are located at homologous regions in human S6 protein. Remarkably, the number, nature, and position of these targeting signals have been conserved, albeit their amino acid sequences have presumably undergone a process of co-evolution with their corresponding rRNAs.     

The nuclear localization signal of mitotic kinesin-like protein Mklp-1: effect on Mklp-1 function during cytokinesis

The mitotic kinesin-like protein (Mklp-1) localizes in the nucleus during interphase due to the presence of nuclear localization signal(s) [NLS(s)] within its sequence. Here, we mapped two NLSs to be 899SRKRRSST906 and 949KRKKP953 in the tail domain of Mklp-1, and showed that ectopic expression of a mutant Mklp-1 without the NLSs leads to cell cycle arrest at cytokinesis, indicating that the NLSs are necessary for Mklp-1 to execute its normal function during cell division. Furthermore, mutation of two serine residues in the first NLS to aspartic acid, which mimics phosphorylation, attenuated its nuclear localization function, suggesting that the function of this NLS might be regulated by phosphorylation.     

Mapping of the nuclear localization signals in open reading frame 2 protein from porcine circovirus type 1

Porcine circovirus type 1 （PCV1） contains two major open reading frames encoding the replication-associated proteins and the major structural capsid （Cap） protein. PCV1 Cap has an N-terminus carrying several potential monopartite or bipartite nuclear localization signals （NLS）. The contribution of these partially overlapping motifs to nuclear importing was identified by expression of mutated PCVI Cap versions fused to enhanced green fluorescent protein （EGFP）. The Cterminus truncated PCV1 Cap-EGFP was localized in nuclei of PK-15 cells similar to the wild-type PCV1 Cap-EGFP, whereas truncation of the N-terminus rendered the fusion protein distributed into cytoplasm, indicating that the nuclear import of PCV1 Cap was efficiently mediated by its N-terminal region. Substitutions of basic residues in stretches 9RR- RR12 or the right part of 25RRPYLAHPAFRNRYRWRRK43 resulted in a diffused distribution of the fusion protein in both nuclei and cytoplasm, indicating that the two NLSs were responsible for restricted nuclear targeting of PCV1 Cap.