Determinants for intracellular sorting of cytoplasmic and nuclear intermediate filaments

The mechanism by which nuclear and cytoplasmic filaments are sorted in vivo was studied by examining which lamin sequences are required to target an otherwise cytoplasmic IF protein, the small neurofilament subunit (NF-L), to the nuclear lamina. By swapping corresponding domains between NF-L and lamin A, nuclear envelope targeting of NF-L was shown to require the presence of the "head" domain, a 42-amino acid sequence unique to lamin rod domains, a nuclear localization signal and the CAAX motif. Replacement of the entire COOH-terminal tail of lamin A with that of NF-L had no discernible effect on nuclear localization of lamin A, provided the substituted NF-L tail contained a NLS and a CAAX motif. This chimeric protein exhibited characteristics more typical of lamin B than that of the parental lamin A. With regard to cytoplasmic assembly properties, substitution of the head domain of lamin A for that of NF-L did not substantially affect the ability of NF-L to coassemble with vimentin in the cytoplasm. In contrast, insertion of a 42-amino acid sequence unique to lamin rod domains into NF-L profoundly affected NF-L coassembly with vimentin indicating that the 42-amino acid insertion in lamins may be important for sorting lamins from cytoplasmic IF proteins.     

Investigation of nuclear architecture with a domain-presenting expression system

We have investigated the topogenic properties of the nucleus by ectopic expression of chimeric proteins consisting of a NLS-modified cytoplasmic filament-forming protein, Xenopus laevis vimentin, and domains of inner nuclear membrane proteins. Whereas the "carrier" without cargo, the NLS-vimentin alone, is deposited in a few nuclear body-type structures (J.M. Bridger, H. Herrmann, C. Münkel, P. Lichter, J. Cell Sci., 111, 1241-1253), the distribution is entirely changed upon coupling with the evolutionarily conserved domain of the lamin B tail, the entire lamin B tail, the amino-terminal nucleoplasmic segment of the lamin B receptor (LBR), and the LEM domain of emerin, respectively. Remarkably, every individual chimeric protein exhibits a completely different distribution. Therefore, we assume that the chimeric parts are specifically recognized by factors engaged in nucleus-specific topogenesis. Thus, the conserved domain of the lamin B tail results in the formation of many small accumulations spread all over the nucleus. The chimera with the complete lamin B tail is deposited in short fibrillar aggregates within the nucleus. It does not mediate the integration of the chimeric protein into the nuclear membrane in cultured cells, indicating that the lamin tail alone is not sufficient to direct the integration of a protein into the lamina in vivo. In contrast, in the nuclear assembly system of Xenopus laevis the recombinant NLS-vimentin-lamin tail protein is concentrated at the nuclear membrane. The LBR chimera is arranged in a "beaded-chain"-type fashion, quite different from the more random deposition of NLS-vimentin alone. To our surprise, the LEM domain of emerin induces the retention of most of the chimeric proteins within the cytoplasm. Hence, it appears to be engaged in a strong cytoplasmic interaction that overrides the nuclear localization signal. Finally, the lamin chimera with the conserved part of the lamin B tail is shown to recruit LBR to the nuclear vimentin bodies and, vice versa, the LBR chimera attracts lamin B in transfected cells, thereby demonstrating their bona fide interaction in vivo.     

Intracellular assembly and sorting of intermediate filament proteins: role of the 42 amino acid lamin insert

Nuclear and cytoplasmic intermediate filament (IF) proteins segregate into two independent cellular networks by mechanisms that are poorly understood. We examined the role of a 42 amino acid (aa) insert unique to vertebrate lamin rod domains in the coassembly of nuclear and cytoplasmic IF proteins by overexpressing chimeric IF proteins in human SW13+ and SW13- cells, which contain and lack endogenous cytoplasmic IF proteins, respectively. The chimeric IF proteins consisted of the rod domain of human nuclear lamin A/C protein fused to the amino and carboxyl-terminal domains of the mouse neurofilament light subunit (NF-L), which contained or lacked the 42 aa insert. Immunofluorescence microscopy was used to follow assembly and targeting of the proteins in cells. Chimeric proteins that lacked the 42 aa insert colocalized with vimentin, whereas those that contained the 42 aa insert did not. When overexpressed in SW13- cells, chimeric proteins containing the 42 aa formed very short or broken cytoplasmic filaments, whereas chimeric proteins that lacked the insert assembled efficiently into long, stable cytoplasmic filaments. To examine the roles of other structural motifs in intracellular targeting, we added two additional sequences to the chimera, a nuclear localization signal (NLS) and a CAAX motif, which are found in nuclear IF proteins. Addition of an NLS alone or an NLS in combination with the CAAX motif to the chimera with the 42 aa insert resulted in cagelike filament that assembled close to the nuclear envelope and nuclear lamina-like targeting, respectively. Our results suggest that the rod domains of eukaryotic nuclear and cytoplasmic IF proteins, which are related to each other, are still compatible upon deletion of the 42 aa insert of coassembly. In addition, NF-L end domains can substitute for the corresponding lamin domains in nuclear lamina targeting.     

Determination of the functional domains involved in nucleolar targeting of nucleolin.

Nucleolin (713 aa), a major nucleolar protein, presents two structural domains: a N-terminus implicated in interaction with chromatin and a C-terminus containing four RNA-binding domains (RRMs) and a glycine/arginine-rich domain mainly involved in pre-rRNA packaging. Furthermore, nucleolin was shown to shuttle between cytoplasm and nucleolus. To get an insight on the nature of nuclear and nucleolar localization signals, a set of nucleolin deletion mutants in fusion with the prokaryotic chloramphenicol acetyltransferase (CAT) were constructed, and the resulting chimeric proteins were recognized by anti-CAT antibodies. First, a nuclear location signal bipartite and composed of two short basic stretches separated by eleven residues was characterized. Deletion of either motifs renders the protein cytoplasmic. Second, by deleting one or more domains implicated in nucleolin association either with DNA, RNA, or proteins, we demonstrated that nucleolar accumulation requires, in addition to the nuclear localization sequence, at least two of the five RRMs in presence or absence of N-terminus. However, in presence of only one RRM the N-terminus allowed a partial targeting of the chimeric protein to the nucleolus.     

A constitutive nucleolar protein identified as a member of the nucleoplasmin family.

Using monoclonal antibodies we have localized a polypeptide, appearing on gel electrophoresis with a Mr of approximately 38,000 and a pI of approximately 5.6, to the granular component of the nucleoli of Xenopus laevis oocytes and a broad range of cells from various species. The protein (NO38) also occurs in certain distinct nucleoplasmic particles but is not detected in ribosomes and other cytoplasmic components. During mitosis NO38-containing material dissociates from the nucleolar organizer region and distributes over the chromosomal surfaces and the perichromosomal cytoplasm; in telophase it re-populates the forming nucleoli. With these antibodies we have isolated from a X. laevis ovary lambda gt11 expression library a cDNA clone encoding a polypeptide which, on one- and two-dimensional gel electrophoresis, co-migrates with authentic NO38. The amino acid sequence deduced from this clone defines a polypeptide of 299 amino acids of mol. wt 33,531 which is characterized by the presence of two domains exceptionally rich in aspartic and glutamic acid, one of them flanked by two putative karyophilic signal heptapeptides. Comparison with other protein sequences shows that NO38 is closely related to the histone-binding, karyophilic protein nucleoplasmin: the first 124 amino acids have 58 amino acid positions in common. Protein NO38 also shows striking homologies to the phosphopeptide region of rat nucleolar protein B23 and the carboxyterminal region of human B23. We propose that protein NO38, which forms distinct homo-oligomers of approximately 7S and Mr of approximately 230,000, is a member of a family of karyophilic proteins, the 'nucleoplasmin family'. It is characterized by its specific association with the nucleolus and might be involved in nuclear accumulation, nucleolar storage and pre-rRNA assembly of ribosomal proteins in a manner similar to that discussed for the role of nucleoplasmin in histone storage and chromatin assembly.     

The alpha-helical rod domain of human lamins A and C contains a chromatin binding site.

We examined regions of human lamins A and C involved in binding to surfaces of mitotic chromosomes. An Escherichia coli expression system was used to produce full-length lamin A and lamin C, and truncated lamins retaining the central alpha-helical rod domain (residues 34-388) but lacking various amounts of the amino-terminal 'head' and carboxy-terminal 'tail' domains. We found that lamin A, lamin C and lamin fragments lacking the head domain and tail sequences distal to residue 431 efficiently assembled into paracrystals and strongly associated with mitotic chromosomes. Furthermore, the lamin rod domain also associated with chromosomes, although efficient chromosome coating required the pH 5-6 conditions needed to assemble the rod into higher order structures. Biochemical assays showed that chromosomes substantially reduced the critical concentration for assembly of lamin polypeptides into pelletable structures. Association of the lamin rod with chromosomes was abolished by pretrypsinization of chromosomes, and was not seen for vimentin (which possesses a similar rod domain). These data demonstrate that the alpha-helical rod of lamins A and C contains a specific chromosome binding site. Hence, the central rod domain of intermediate filament proteins can be involved in interactions with other cellular structures as well as in filament assembly.     

Human autoantibodies against a nucleolar protein

Autoimmune sera showing prominent immunofluorescence in nucleolus were selected and analysed by immunoblotting techniques. Immunoblots using a nucleolar extract as antigen source revealed sera recognizing a 38 kDa nucleolar protein. Low concentration of Actinomycin D, which inhibits the ribosomal RNA synthesis, caused a loss of fluorescence. This suggests that the nucleolar antigen may be associated with the assembly of packaging of the ribosomes. The present nucleolar antigen has properties similar to the previously described nucleolar phosphoprotein B23 of rat cells and the recently described nucleolar protein NO38 of mouse and Xenopus cells.     

Isolation and sequencing of NOP1. A yeast gene encoding a nucleolar protein homologous to a human autoimmune antigen

We have identified the gene for the yeast nucleolar protein p38 and deduced the primary structure of p38 from its sequence. We propose the name NOP1 (nucleolar protein 1) for this gene. NOP1 encodes a 327 amino acid protein of 34,470 daltons and is flanked by potential promoter and polyadenylation sequences. Blot analyses indicate that the mRNA transcribed from NOP1 is approximately 1.3 kilobases in size and that there is one NOP1 gene per haploid genome. The amino-terminal sequence of p38 is homologous with the 31 known amino-terminal residues of the autoimmune antigen fibrillarin, confirming the previously observed similarity between p38 and this mammalian nucleolar protein. Consistent with this, p38 cross-reacts with serum from a patient with the autoimmune disease scleroderma. A putative nuclear localization signal can be identified in p38. Interestingly, a repetitive amino acid sequence motif begins near the amino terminus of p38. This motif is approximately 80 residues long, is rich in glycine and arginine, and shows striking sequence homology to mammalian nucleolins and certain nucleic acid binding proteins.     

Characterization of dominant and recessive assembly-defective mutations in mouse neurofilament NF-M

We have generated a set of amino- and carboxy-terminal deletions of the neurofilament NF-M gene and determined the molecular consequences of forced expression of these mutant constructs in mouse fibroblasts. To follow the expression of mutant NF-M subunits in transfected cells, a 12 amino acid epitope (from the human c-myc protein) was expressed at the carboxy terminus of each mutant. We show that NF-M molecules missing up to 90 or 70% of the nonhelical carboxy-terminal tail or amino-terminal head domains, respectively, incorporate readily into an intermediate filament network comprised either of vimentin or NF-L, whereas deletions into either the amino- or carboxy-terminal alpha- helical rod region generate assembly-incompetent polypeptides. Carboxy- terminal deletions into the rod domain invariably yield dominant mutants which rapidly disrupt the array of filaments comprised of NF-L or vimentin. Accumulation of these mutant NF-M subunits disrupts vimentin filament arrays even when present at approximately 1% the level of the wild-type subunits. In contrast, the amino-terminal deletions into the rod produce pseudo-recessive mutants that perturb the wild-type NF-L or vimentin arrays only modestly. The inability of such amino-terminal mutants to disrupt wild-type subunits defines a region near the amino-terminal alpha-helical rod domain (residues 75- 126) that is required for the earliest steps in filament assembly.     

Protein B23 is an important human factor for the nucleolar localization of the human immunodeficiency virus protein Tat.

Nucleolar shuttle protein B23 was found to bind to human immunodeficiency virus protein Tat, and this binding required the nucleolar localization motif of Tat. A fusion protein containing the B23 binding domain and beta-galactosidase caused mislocalization of Tat to the cytoplasm and inhibited the transactivation activity of Tat. These data suggest that B23 is a human factor necessary for the nucleolar localization of Tat.     

Quantitative analysis of localization and nuclear aggregate formation induced by GFP-lamin A mutant proteins in living HeLa cells

Although A-type lamins are ubiquitously expressed, their role in the tissue-specificity of human laminopathies remains enigmatic. In this study, we generate a series of transfection constructs encoding missense lamin A mutant proteins fused to green fluorescent protein and investigate their subnuclear localization using quantitative live cell imaging. The mutant constructs used included the laminopathy-inducing lamin A rod domain mutants N195K, E358K, M371K, R386K, the tail domain mutants G465D, R482L, and R527P, and the Hutchinson-Gilford progeria syndrome-causing deletion mutant, progerin (LaA delta50). All mutant derivatives induced nuclear aggregates, except for progerin, which caused a more lobulated phenotype of the nucleus. Quantitative analysis revealed that the frequency of nuclear aggregate formation was significantly higher (two to four times) for the mutants compared to the wild type, although the level of lamin fusion proteins within nuclear aggregates was not. The distribution of endogenous A-type lamins was altered by overexpression of the lamin A mutants, coexpression experiments revealing that aberrant localization of the N195K and R386K mutants had no effect on the subnuclear distribution of histones H2A or H2B, or on nuclear accumulation of H2A overexpressed as a DsRed2 fusion protein. The GFP-lamin fusion protein-expressing constructs will have important applications in the future, enabling live cell imaging of nuclear processes involving lamins and how this may relate to the pathogenesis of laminopathies.     

Identification and characterization of a yeast nucleolar protein that is similar to a rat liver nucleolar protein

We have produced monoclonal antibodies against purified nuclei from the yeast Saccharomyces cerevisiae and have characterized three different antibodies that recognize a protein with an apparent molecular weight of 38,000, termed p38. Subcellular fractionation shows that virtually all of p38 occurs in the nuclear fraction. High concentrations of salt (1 M) or urea (6 M) effectively solubilize p38 from a nuclear envelope fraction prepared by digestion of nuclei with DNase. Indirect immunofluorescence demonstrates a crescent shaped distribution of p38 at the inner periphery of the nucleus, with p38 extending between dividing pairs of cells during (closed) mitosis. Postembedding immunogold electron microscopy shows decoration of the densely stained "crescent" region of the yeast nucleus, confirming the localization of p38 to the nucleolus. One of the monoclonals, D77, cross reacts on immunoblots with a single protein of molecular weight 37,000 from purified rat liver nuclei. Indirect immunofluorescence localizes this protein to the nucleolus, and shows that it is dispersed throughout the cell during mitosis. The yeast and rat liver nucleolar proteins behave similarly when electrophoresed in two dimensions, and appear to have basic pI values. Analysis of immunological cross-reactivity using D77, and antibodies specific for nucleolar proteins from other sources, suggests that the rat liver protein is fibrillarin, and demonstrates that p38 shares epitopes with fibrillarin, as well as with other vertebrate nucleolar proteins.     

The role of the head and tail domain in lamin structure and assembly: analysis of bacterially expressed chicken lamin A and truncated B2 lamins.

Nuclear lamins like cytoplasmic intermediate filament proteins exhibit a characteristic tripartite domain structure with a segmented alpha-helical rod domain flanked by an N-terminal head and a C-terminal tail domain. To examine the influence of the head and tail domains on the structure and assembly properties of nuclear lamins, we have engineered "headless," "tailless," and "rod" chicken lamin B2 cDNAs and expressed them in Escherichia coli. A full-length chicken lamin A cDNA was also expressed in E. coli, and the recombinant protein compared with the structure and assembly properties of full-length chicken lamin B2 (E. Heitlinger et al. (1991) J. Cell Biol. 113, 485-495). As with lamin B2, at their first level of structural organization, lamin A and the headless lamin B2 formed myosin-like dimers consisting of a 51- to 52-nm-long tail flanked by two globular heads at one end. Similarly, the tailless and rod lamin B2 fragments formed tropomyosin-like dimers consisting of a 51 to 52-nm-long rod. In contrast to the lateral mode of association of cytoplasmic IF dimers into four-chain tetramers, at their second level of structural organization, lamin A dimers, just as lamin B2 dimers (E. Heitlinger et al. (1991) J. Cell Biol. 113, 485-495), associated longitudinally to form polar head-to-tail polymers. Whereas dimers made of the truncated B2 headless and rod lamins had lost their propensity to associate head-to-tail, tailless lamin B2 dimers revealed an enhanced head-to-tail association. Finally, at their third level of structural organization, rather than assembling into stable 10-nm filaments, both lamin A and the three truncated B2 lamins formed paracrystalline arrays exhibiting distinct transverse banding patterns with axial repeats of either 24 or 48-49 nm depending on the species.     

Novel nucleolar protein, midnolin, is expressed in the mesencephalon during mouse development

Using the gene trap method and the selection of embryonic stem cells in vitro, we have identified several novel genes involved in mouse development. The detailed analysis of one of these, named midnolin (midbrain nucleolar protein), is reported here. Expression of the midnolin gene is developmentally regulated: it is strongly expressed at the mesencephalon (midbrain) of the embryo in day 12.5 (E12.5) mice. The midnolin encodes a protein of 508 amino acids (aa), which contains a Ubiquitin-like domain. The intracellular distribution of the midnolin was studied by using midnolin-green fluorescent protein (GFP) fusion proteins. Midnolin was found to be localized in the nucleus and nucleolus, but not in the cytoplasm. The nucleolar localization signal was determined to be a 28aa peptide (440-QQKRLRRKARRDARGPYHWTPSRKAGRS-467) located at the C-terminal region of the midnolin. Our results suggest that midnolin is involved in regulation of genes related to neurogenesis in the nucleolus.     

Both lamin A and lamin C mutations cause lamina instability as well as loss of internal nuclear lamin organization

We have applied the fluorescence loss of intensity after photobleaching (FLIP) technique to study the molecular dynamics and organization of nuclear lamin proteins in cell lines stably transfected with green fluorescent protein (GFP)-tagged A-type lamin cDNA. Normal lamin A and C proteins show abundant decoration of the inner layer of the nuclear membrane, the nuclear lamina, and a generally diffuse localization in the nuclear interior. Bleaching studies revealed that, while the GFP-tagged lamins in the lamina were virtually immobile, the intranuclear fraction of these molecules was partially mobile. Intranuclear lamin C was significantly more mobile than intranuclear lamina A. In search of a structural cause for the variety of inherited diseases caused by A-type lamin mutations, we have studied the molecular organization of GFP-tagged lamin A and lamin C mutants R453W and R386K, found in Emery-Dreifuss muscular dystrophy (EDMD), and lamin A and lamin C mutant R482W, found in patients with Dunnigan-type familial partial lipodystrophy (FPLD). In all mutants, a prominent increase in lamin mobility was observed, indicating loss of structural stability of lamin polymers, both at the perinuclear lamina and in the intranuclear lamin organization. While the lamin rod domain mutant showed overall increased mobility, the tail domain mutants showed mainly intranuclear destabilization, possibly as a result of loss of interaction with chromatin. Decreased stability of lamin mutant polymers was confirmed by flow cytometric analyses and immunoblotting of nuclear extracts. Our findings suggest a loss of function of A-type lamin mutant proteins in the organization of intranuclear chromatin and predict the loss of gene regulatory function in laminopathies.     

In vitro assembly properties of mutant and chimeric intermediate filament proteins: insight into the function of sequences in the rod and end domains of IF

The factors and mechanisms regulating assembly of intermediate filament (IF) proteins to produce filaments with their characteristic 10 nm diameter are not fully understood. All IF proteins contain a central rod domain flanked by variable head and tail domains. To elucidate the role that different domains of IF proteins play in filament assembly, we used negative staining and electron microscopy (EM) to study the in vitro assembly properties of purified bacterially expressed IF proteins, in which specific domains of the proteins were either mutated or swapped between a cytoplasmic (mouse neurofilament-light (NF-L) subunit) and nuclear intermediate filament protein (human lamin A). Our results indicate that filament formation is profoundly influenced by the composition of the assembly buffer. Wild type (wt) mouse NF-L formed 10 nm filaments in assembly buffer containing 175 mM NaCl, whereas a mutant deleted of 18 NH2-terminal amino acids failed to assemble under similar conditions. Instead, the mutant assembled efficiently in buffers containing CaCl2 > or = 6 mM forming filaments that were 10 times longer than those formed by wt NF-L, although their diameter was significantly smaller (6-7 nm). These results suggest that the 18 NH2-terminal sequence of NF-L might serve two functions, to inhibit filament elongation and to promote lateral association of NF-L subunits. We also demonstrate that lengthening of the NF-L rod domain, by inserting a 42 aa sequence unique to nuclear IF proteins, does not compromise filament assembly in any noticeable way. Our results suggests that the known inability of nuclear lamin proteins to assemble into 10 nm filaments in vitro cannot derive solely from their longer rod domain. Finally, we demonstrate that the head domain of lamin A can substitute for that of NF-L in filament assembly, whereas substitution of both the head and tail domains of lamins for those of NF-L compromises assembly. Therefore, the effect of lamin A "tail" domain alone, or the synergistic effect of lamin "head" and the "tail" domains together, interferes with assembly into 10-nm filaments.     

Tryptophans 286 and 288 in the C-terminal region of protein B23.1 are important for its nucleolar localization

Nucleolar protein B23 can shuttle between the nucleolus and cytoplasm. However, the mechanism involved in the protein moving and staying in the nucleolus is not fully understood. To identify the nucleolar localization signal sequence of protein B23, we examined the subnuclear location of B23.1 mutant proteins fused with green fluorescent protein in HeLa cells. The results suggested that the two C-terminal tryptophan residues (Trp-286 and Trp-288) of protein B23.1 were important in this phenomenon.     

Epitope distribution and immunochemical characterization of nucleolar phosphoprotein C23 using ten monoclonal antibodies

Summary Extractable nucleolar proteins from HeLa cells were used as a source of antigen to immunize mice for monoclonal antibody (MAb) production. Ten of the resulting MAbs shown to identify nucleolar phosphoprotein (110 kD/pI 5.5) were purified and used in immunochemical studies to further characterize protein C23. All ten MAbs showed nucleolar localization by indirect immunofluorescence; one antibody (FR2) also showed some nucleoplasmic localization that was attributed to a shared epitope between protein C23 and a 72 kD nuclear/nucleolar antigen. Reciprocal antibody cross blocking studies indicated that the ten MAbs identified nine distinct epitopes on protein C23. Interestingly, seven of the nine epitopes were shown by immunofluorescence and competitive ELISA studies to be species related. Immunostained patterns of exponentially growing HeLa cells suggest that protein C23 exists in vivo solely as a 110 kD peptide. However, protein C23 was subject to rapid degradation into a number of proteolytic fragments upon extraction or storage of isolated nucleoli. The failure to find protein C23 related peptides with molecular sizes less than 110 kD in exponentially growing cells and the lack of cytoplasmic localization of any of the ten MAbs suggests that protein C23 is not a prepro-protein processed in vivo to form ribosomal proteins as previously suggested (1).