Two wobble-splicing events affect ING4 protein subnuclear localization and degradation

ING4 (inhibitor of growth 4) is a candidate tumor suppressor gene that is implicated as a repressor of cell growth, angiogenesis, cell spreading and cell migration and can suppress loss of contact inhibition in vitro. Another group and we identified four wobble-splicing isoforms of ING4 generated by alternative splicing at two tandem splice sites, GC(N)₇GT and NAGNAG, which caused canonical (GT-AG) and non-canonical (GC-AG) splice site wobbling selection. Expression of the four ING4 wobble-splicing isoforms did not vary significantly in any of the cell lines examined. Here we show that ING4_v1 is translocated to the nucleolus, indicating that ING4 contains an intrinsic nucleolar localization signal. We further demonstrate that the subcellular localization of ING4 is modulated by two wobble-splicing events at the exon 4-5 boundary, causing displacement from the nucleolus to the nucleus. We also observed that ING4 is degraded through the ubiquitin-proteasome pathway and that it is subjected to N-terminal ubiquitination. We demonstrate that nucleolar accumulation of ING4 prolongs its half-life, but lack of nucleolar targeting potentially increases ING4 degradation. Taken together, our data suggest that the two wobble-splicing events at the exon 4-5 boundary influence subnuclear localization and degradation of ING4.     

Nuclear transport of Ras-associated tumor suppressor proteins: different transport receptor binding specificities for arginine-rich nuclear targeting signals

Ras proteins regulate a wide range of biological processes by interacting with a variety of effector proteins. In addition to the known role in tumorigensis, the activated form of Ras exhibits growth-inhibitory effects by unknown mechanisms. Several Ras effector proteins identified as mediators of apoptosis and cell-cycle arrest also exhibit properties normally associated with tumor suppressor proteins. Here, we show that Ras effector RASSF5/NORE-1 binds strongly to K-Ras but weakly to both N-Ras and H-Ras. RASSF5 was found to localize both in the nucleus and the nucleolus in contrast to other Ras effector proteins, RASSF1C and RASSF2, which are localized in the nucleus and excluded from nucleolus. A 50 amino acid residue transferable arginine-rich nucleolar localization signal (NoLS) identified in RASSF5 is capable of interacting with importin-beta and transporting the cargo into the nucleolus. Surprisingly, similar arginine-rich signals identified in RASSF1C and RASSF2 interact with importin-alpha and transport the heterologous cytoplasmic proteins to the nucleus. Interestingly, mutation of arginine residues within these nuclear targeting signals prevented interaction of Ras effector proteins with respective transport receptors and abolished their nuclear translocation. These results provide evidence for the first time that arginine-rich signals are able to recognize different nuclear import receptors and transport the RASSF proteins into distinct sub-cellular compartments. In addition, our data suggest that the nuclear localization of RASSF5 is critical for its cell growth control activity. Together, these data suggest that the transport of Ras effector superfamily proteins into the nucleus/nucleolus may play a vital role in modulating Ras-mediated cell proliferation during tumorigenesis.     

灰盖拟鬼伞核定位蛋白重组表达系统的构建

为构建灰盖拟鬼伞Coprinopsis cinerea的核定位蛋白重组表达系统,本研究通过蛋白序列比对和信息学分析,预测了灰盖拟鬼伞组蛋白H2B的核定位序列,构建了融合组蛋白H2B核定位序列的绿色荧光蛋白(green fluorescent protein,GFP)重组表达载体,将该载体转入灰盖拟鬼伞AmutBmut菌...     

Characterization of the nuclear localization signal of the mouse TET3 protein

DNA demethylation is associated with gene activation and is mediated by a family of ten-eleven translocation (TET) dioxygenase. The TET3 protein is a 1668-amino-acid DNA demethylase that is predicted to possess five nuclear localization signals (NLSs). In this paper, we used a series of green fluorescent protein-tagged and mutation constructs to identify a conserved NLS (KKRK) embedded between amino acid 1615 and 1618 of mouse TET3. The KKRK sequence facilitates the cytoplasmic protein’s translocation into the nucleus. Additionally TET3 may be imported into the nucleus by importin-α and importin-β.     

Characterization of nuclear localization and nuclear export signals of yeast actin-binding protein Pan1

Pan1 is an actin patch-associated protein involved in endocytosis. Our studies revealed that in oleate-grown cells Pan1 is located in the nucleus as well as in patches. One of three putative nuclear localization signals (NLS) of Pan1, NLS2, directed beta-galactosidase (beta-gal) to the nucleus. However, GFP-Pan1(886-1219), containing NLS2, was found in the cytoplasm indicating that it may contain a nuclear export signal (NES). A putative Pan1 NES, overlapping with NLS3, re-addressed NLS(H2B)-NES/NLS3-beta-gal from the nucleus to the cytoplasm. Inactivation of the NES allowed NLS3 to be effective. Thus, Pan1 contains functional NLSs and a NES and appears to shuttle in certain circumstances.     

Glycine-rich region regulates cysteine-rich protein 1 binding to actin cytoskeleton

Cysteine-rich protein 1 (CRP1) has a unique structure with two well separated LIM domains, each followed by a glycine-rich region. Although CRP1 has been shown to interact with actin-binding proteins and actin filaments, the mechanism regulating localization to the actin cytoskeleton in cells is not clear. Experiments using truncated forms showed that the first LIM domain and glycine-rich region are necessary for CRP1 bundling of actin filaments and localization to the actin cytoskeleton. Furthermore, domain swapping experiments replacing the first glycine-rich region with the second resulted in the loss of CRP1 bundling activity and localization to the actin cytoskeleton, identifying seven critical amino acid residues. These results highlight the importance of the first glycine-rich region for CRP1 bundling activity and localization to the actin cytoskeleton. In addition, this work identifies the first LIM domain and glycine-rich region as a distinct actin filament bundling module.     

Glucose-induced dissociation of glucokinase from its regulatory protein in the nucleus of hepatocytes prior to nuclear export

The glucose phosphorylating enzyme glucokinase regulates glucose metabolism in the liver. Glucokinase activity is modulated by a liver-specific competitive inhibitor, the glucokinase regulatory protein (GRP), which mediates sequestration of glucokinase to the nucleus at low glucose concentrations. However, the mechanism of glucokinase nuclear export is not fully understood. In this study we investigated the dynamics of glucose-dependent interaction and translocation of glucokinase and GRP in primary hepatocytes using fluorescence resonance energy transfer, selective photoconversion and fluorescence recovery after photobleaching. The formation of the glucokinase:GRP complex in the nucleus of primary hepatocytes at 5 mmol/l glucose was significantly reduced after a 2 h incubation at 20 mmol/l glucose. The GRP was predominantly localized in the nucleus, but a mobile fraction moved between the nucleus and the cytoplasm. The glucose concentration only marginally affected GRP shuttling. In contrast, the nuclear export rate of glucokinase was significantly higher at 20 than at 5 mmol/l glucose. Thus, glucose was proven to be the driving-force for nuclear export of glucokinase in hepatocytes. Using the FLII12Pglu-700μ-δ6 glucose nanosensor it could be shown that in hepatocytes the kinetics of nuclear glucose influx, metabolism or efflux were significantly faster compared to insulin-secreting cells. The rapid equilibration kinetics of glucose flux into the nucleus facilitates dissociation of the glucokinase:GRP complex and also nuclear glucose metabolism by free glucokinase enzyme. In conclusion, we could show that a rise of glucose in the nucleus of hepatocytes releases active glucokinase from the glucokinase:GRP complex and promotes the subsequent nuclear export of glucokinase.     

Roles of the intramolecular regions of FE65 in its trans-accumulation and in p53 stabilization in the nuclear matrix of osmotically stressed cells

The neural adaptor protein FE65 interacts with the amyloid β-protein precursor (APP). In osmotically stressed cells, the membrane APP-tethered FE65 is released into the cytoplasm and translocates to the nuclear matrix, where it stabilizes p53 via a non-canonical pathway. In this study, we found that the second phosphotyrosine interaction domain (PI2) of FE65 mediated its trans-accumulation in the nuclear matrix of osmotically stressed cells. The carboxyl-terminal half of FE65, which contains the PI2 domain, failed to stabilize p53, suggesting that the amino-terminal half of the protein plays an important role in the stabilization of p53 in osmotically stressed cells.     

N-terminal glycine-specific protein conjugation catalyzed by microbial transglutaminase

Here, we report the N-terminal glycine (Gly) residue of a target protein can be a candidate primary amine for site-specific protein conjugation catalyzed by microbial transglutaminase (MTG) from Streptomyces mobaraensis. Gly5-enhanced green fluorescent protein (EGFP) (EGFP with five additional Gly residues at its N-terminus) was cross-linked with Myc-dihydrofolate reductase (DHFR) (DHFR with the myc epitope sequence at its N-terminus) to yield DHFR-EGFP heterodimers. The reactivities of additional peptidyl linkers were investigated and the results obtained suggested that at least three additional Gly residues at the N-terminus were required to yield the EGFP-DHFR heterodimeric form. Site-directed mutagenesis analysis revealed marked preference of MTG for amino acids adjacent to the N-terminal Gly residue involved in the protein conjugation. In addition, peptide-protein conjugation was demonstrated by MTG-catalyzed N-terminal Gly-specific modification of a target protein with the myc epitope peptide.     

Alpha helical structures in the leader sequence of human GLUD2 glutamate dehydrogenase responsible for mitochondrial import

Human glutamate dehydrogenase (hGDH) exists in two highly homologous isoforms with a distinct regulatory and tissue expression profile: a housekeeping hGDH1 isoprotein encoded by the GLUD1 gene and an hGDH2 isoenzyme encoded by the GLUD2 gene. There is evidence that both isoenzymes are synthesized as pro-enzymes containing a 53 amino acid long N-terminal leader peptide that is cleaved upon translocation into the mitochondria. However, this GDH signal peptide is substantially larger than that of most nuclear DNA-encoded mitochondrial proteins, the leader sequence of which typically contains 17-35 amino acids and they often form a single amphipathic α-helix. To decode the structural elements that are essential for the mitochondrial targeting of human GDHs, we performed secondary structure analyses of their leader sequence. These analyses predicted, with 82% accuracy, that both leader peptides are positively charged and that they form two to three α-helices, separated by intermediate loops. The first α-helix of hGDH2 is strongly amphipathic, displaying both a positively charged surface and a hydrophobic plane. We then constructed GLUD2-EGFP deletion mutants and used them to transfect three mammalian cell lines (HEK293, COS 7 and SHSY-5Y). Confocal laser scanning microscopy, following co-transfection with pDsRed2-Mito mitochondrial targeting vector, revealed that deletion of the entire leader sequence prevented the enzyme from entering the mitochondria, resulting in its retention in the cytoplasm. Deletion of the first strongly amphipathic α-helix only was also sufficient to prevent the mitochondrial localization of the truncated protein. Moreover, truncated leader sequences, retaining the second and/or the third putative α-helix, failed to restore the mitochondrial import of hGDH2. As such, the first N-terminal alpha helical structure is crucial for the mitochondrial import of hGDH2 and these findings may have implications in understanding the evolutionary mechanisms that led to the large mitochondrial targeting signals of human GDHs.     

A novel lysine-rich domain and GTP binding motifs regulate the nucleolar retention of human guanine nucleotide binding protein, GNL3L

A variety of G-proteins and GTPases are known to be involved in nucleolar function. We describe here a new evolutionarily conserved putative human GTPase, guanine nucleotide binding protein-like 3-like (GNL3L). Genes encoding proteins related to GNL3L are present in bacteria and yeast to metazoa and suggests its critical role in development. Conserved domain search analysis revealed that the GNL3L contains a circularly permuted G-motif described by a G5-G4-G1-G2-G3 pattern similar to the HSR1/MMR1 GTP-binding protein subfamily. Highly conserved and critical residues were identified from a three-dimensional structural model obtained for GNL3L using the crystal structure of an Ylqf GTPase from Bacillus subtilis. We demonstrate here that GNL3L is transported into the nucleolus by a novel lysine-rich nucleolar localization signal (NoLS) residing within 1-50 amino acid residues. NoLS identified here is necessary and sufficient to target the heterologous proteins to the nucleolus. We show for the first time that the lysine-rich targeting signal interacts with the nuclear transport receptor, importin-beta and transports GNL3L into the nucleolus. Interestingly, depletion of intracellular GTP blocks GNL3L accumulation into the nucleolar compartment. Furthermore, mutations within the G-domains alter the GTP binding ability of GNL3L and abrogate wild-type nucleolar retention even in the presence of functional NoLS, suggesting that the efficient nucleolar retention of GNL3L involves activities of both basic NoLS and GTP-binding domains. Collectively, these data suggest that GNL3L is composed of distinct modules, each of which plays a specific role in molecular interactions for its nucleolar retention and subsequent function(s) within the nucleolus.     

Extremely low polymerizability of a highly-divergent Chlamydomonas actin (NAP)

Novel actin-like protein (NAP) is a highly divergent actin expressed in Chlamydomonas. With its low sequence similarity, it is uncertain whether NAP can polymerize into filaments. Here I assessed it by ectopically expressing enhanced green fluorescent protein-tagged NAP (EGFP-NAP) in cultured cells. EGFP-NAP was excluded from stress fibres but partially co-localized with endogenous actin in the cell periphery. In fluorescence recovery after photobleaching experiment, turnover rate of EGFP-NAP was similar to the estimated diffusion rate of monomeric actin. Therefore, EGFP-NAP likely accumulates by diffusion. These findings suggest that NAP has extremely poor ability to polymerize.     

Characterization of a nuclear compartment shared by nuclear bodies applying ectopic protein expression and correlative light and electron microscopy

To investigate the accessibility of interphase nuclei for nuclear body-sized particles, we analyzed in cultured cells from human origin by correlative fluorescence and electron microscopy (EM) the bundle-formation of Xenopus-vimentin targeted to the nucleus via a nuclear localization signal (NLS). Moreover, we investigated the spatial relationship of speckles, Cajal bodies, and crystalline particles formed by Mx1 fused to yellow fluorescent protein (YFP), with respect to these bundle arrays. At 37 degrees C, the nucleus-targeted, temperature-sensitive Xenopus vimentin was deposited in focal accumulations. Upon shift to 28 degrees C, polymerization was induced and filament arrays became visible. Within 2 h after temperature shift, arrays were found to be composed of filaments loosely embedded in the nucleoplasm. The filaments were restricted to limited areas of the nucleus between focal accumulations. Upon incubation at 28 degrees C for several hours, NLS vimentin filaments formed bundles looping throughout the nuclei. Speckles and Cajal bodies frequently localized in direct neighborhood to vimentin bundles. Similarly, small crystalline particles formed by YFP-tagged Mx1 also located next to vimentin bundles. Taking into account that nuclear targeted vimentin locates in the interchromosomal domain (ICD), we conclude that nuclear body-sized particles share a common nuclear space which is controlled by higher order chromatin organization.