Mobility Within the Nucleus and Neighboring Cytosol Is a Key Feature of Prothymosin-α

Abstract

Prothymosin α is a small, unfolded, negatively charged, poorly antigenic mammalian protein with a potent nuclear localization signal. Although it is apparently essential for growth, its precise function is unknown. We examined the location and behavior of the protein bearing different epitope tags using in situ immunolocalization in COS-1 and NIH3T3 cells. Tagged prothymosin α appeared to be punctate and widely dispersed throughout the nucleus, with the exception of the nucleolus. A tiny cytoplasmic component, which persisted in the presence of cycloheximide and actinomycin D during interphase, became pronounced immediately before, during, and after mitosis. When nuclear uptake was abrogated, small tagged prothymosin α molecules, but not prothymosin α fused to β-galactosidase, accumulated significantly in the cytoplasm. Tagged prothymosin α shared domains with mobile proteins such as Ran, transportin, and karyopherin β, which also traverse the nuclear membrane, and co-localized with active RNA polymerase II. Mild digitonin treatment resulted in nuclei devoid of prothymosin α. The data do not support tight binding to any nuclear component. Therefore, we propose that prothymosin α is a highly diffusible bolus of salt and infer that it facilitates movement of charged molecules in highly charged environments within and near the nucleus.

Keywords

prothymosin α nucleus mitosis charged molecules

The function of prothymosin α remains unclear despite many attempts to determine both its binding partners and its cellular location. Although the levels of the protein and its mRNA unquestionably increase in proliferating mammalian cells (Eschenfeldt and Berger 1986; Gómez-Márquez et al. 1989; Bustelo et al. 1991; Tsitsiloni et al. 1993), precisely what prothymosin α does, where it does it, and what other macromolecules participate are questions with multiple unsatisfactory answers. The difficulties stem from three attributes: (a) Prothymosin α is extremely acidic, with a net negative charge of ∼40 units (Eschenfeldt and Berger 1986; Goodall et al. 1986; Frangou-Lazaridis et al. 1988; Panneerselvam et al. 1988; Schmidt and Werner 1991). The large number of contiguous acidic amino acids in a protein of only ∼12 kD (Haritos et al. 1985), an unfolded conformation (Gast et al. 1995), and an abundance approaching that of histone cores (Palvimo and Linnala–Kankkunen 1990; Sburlati et al. 1990, 1993) make prothymosin α an easy target for basic proteins in binding studies performed in vitro. The relevance of binding data is, accordingly, problematic. (b) Antibodies raised to prothymosin α or peptides derived from it are invariably low in titer and broad in specificity. Such antibodies, particularly those directed against the N-terminus, crossreact with materials from crabs, insects, protozoans, fungi, and bacteria (Oates and Erdos 1989). However, the fully sequenced yeast genome does not code for prothymosin α or related proteins, and homologous proteins in bacteria and homologous genes in a broad swath of the evolutionary spectrum up to and including reptiles have not been forthcoming (Trumbore et al. 1998). Clearly, studies with antibodies able to crossreact with highly divergent antigens should be approached with caution. Finally, (c) prothymosin α does not have an activity that can be used in vivo or in vitro to assess its functional integrity.

The published binding partners of prothymosin α in vitro include histone H1 and core histones (Papmarcaki and Tsolas 1994; Díaz-Jullien et al. 1996; Karetsou et al. 1998), which help to package DNA and regulate transcription, the rev protein of HIV and the rex protein of HTLV (Kubota et al. 1995), which are involved in transport of RNA out of the nucleus, and a small cytoplasmic RNA, reported by Vartapetian and co-workers (1988) to be covalently attached to prothymosin α (Makarova et al. 1989). The aforementioned group assumed that a protein remaining in the aqueous phase after extraction with phenol must be nucleic acid-bound; later, it was shown that prothymosin α itself partitions to the aqueous phase and behaves chemically much like an RNA (Sburlati et al. 1990). It is now fairly certain that a stable N-terminal acetyl group (Goldstein et al. 1977; Michalewsky et al. 1983; Haritos et al. 1984; Sburlati et al. 1993) and several highly labile glutamyl phosphate groups (Trumbore et al. 1997), which are not retained in vitro, are the only physiological post-translational modifications. In addition, one study claiming that prothymosin α stimulates the phosphorylation of translation factor eEF2 during mitosis (Vega et al. 1998) has been rigorously challenged (Enkemann et al. 1999). From these examples, it should be evident that compelling associations have yet to emerge.

During its 15-year history, prothymosin α has been localized to several sectors both inside and outside the cell. As the name suggests, prothymosin α was once believed to be a precursor of a putative thymic hormone, thymosin α1 (Haritos et al. 1984), now known to be an artifactual proteolytic degradation product consisting of amino acids 1–28 (reviewed in Szabo and Weksler 1992). Soon afterwards, it was scrutinized as the secreted commodity itself. In this guise, investigators looked for and found tiny amounts of intact prothymosin α in blood (Panneerselvam et al. 1987) but could not identify a means of egress from cells or tissues. When the gene for prothymosin α was isolated, the absence of a signal peptide, together with the presence of a consensus nuclear localization signal, strongly implied a nuclear destination (Eschenfeldt and Berger 1986; Goodall et al. 1986; Eschenfeldt et al. 1989; Watts et al. 1990). This view was supported by several avenues of biochemical research. Prothymosin α or a derived peptide bearing the consensus bipartite basic cluster of amino acids near the carboxyl terminus was able to drag covalently-attached non-nuclear proteins into the nucleus (Manrow et al. 1991). Centrifugal enucleation of cytochalasin-treated COS-1 cells, a procedure that generates nuclei sealed within a fragment of plasma membrane, caused the entirety of prothymosin α fused to β-galactosidase to remain nuclear (Manrow et al. 1991). Xenopus oocytes injected with bovine prothymosin α labeled with Bolton Hunter reagent on the e-NH₂ groups of lysine residues transported a significant fraction to the nucleus (Watts et al. 1990). Nevertheless, in situ immunolocalization studies returned a mixed verdict. A mouse monoclonal antibody to thymosin α₁ recognized material in the cytoplasm of epithelial cells in the subcapsullary and medullary zones of the thymus and, at the electron microscopic level, stained large and small cytoplasmic vacuoles (Auger et al. 1987). In contrast, polyvalent antibodies to thymosin α₁ highlighted immunoreactive material in the nuclei of bile duct cells but not hepatocytes (Fraga et al. 1993), and distinguished “thymosin immunoreactive peptides” primarily in the nuclei of IEC-6 cells, a line derived from the small intestine of the rat (Conteas et al. 1990). In disrupted cells, prothymosin α was almost invariably cytoplasmic (Tsitsiloni et al. 1989; Sburlati et al. 1990). Therefore, these studies, too, failed to dispel the uncertainty surrounding the in vivo location of prothymosin α.

It is now widely accepted that prothymosin α is a nuclear protein (Watts et al. 1989, 1990; Clinton et al. 1991; Manrow et al. 1991). However, observations of a cytoplasmic component are too widespread to be summarily dismissed. Our present study sought to resolve the ambiguities by defining the cellular position of prothymosin α under a wide variety of metabolic conditions. In this endeavor, we chose to employ the protein tagged with several different epitopes and a series of antibodies that recognize them. Antibodies to prothymosin α itself were avoided because without supportive information they are suspect and unsuitable for work in situ. We have also investigated potential neighbors of prothymosin α. Our experiments suggest that the protein is overwhelmingly nuclear, highly mobile, intermittently cytoplasmic, and continuously pumped into the nucleus. It is apparently not tightly tethered to any macromolecular species either in the nucleus or in the cytoplasm. To explain the data, we propose that prothymosin α is a freely diffusible bolus of concentrated salt that facilitates movement of charged molecules in highly charged environments near and within the nucleus.

Materials and Methods

Cell Culture

Origin-defective SV40-transformed green monkey kidney cells (COS-1) were grown in Dulbecco's modified Eagle's medium (DMEM) from Biofluids (catalog #172; Rockville, MD) supplemented with 10% heat-inactivated fetal calf serum (Hyclone; Logan, UT), 2 mM l-glutamine, 90 U/ml of penicillin, 90 μg/ml of streptomycin, and 0.22 μg/ml of amphotericin B (Life Technologies; Gaithersburg, MD). NIH3T3 cells were cultured similarly except for the substitution of calf serum for fetal calf serum. All cultures were incubated in a humidified environment containing 5% CO₂ at 37C unless noted otherwise. For inhibitor studies, cells were cultured in the presence of 5 μg/ml of actinomycin D or 100 μg/ml of cycloheximide. Stock solutions of actinomycin D were prepared in ethanol; accordingly, samples and controls contained 0.25% (v/v) solvent.

Plasmids and Transfection Procedures

Plasmid pBC12BIProG 6-his, which codes for prothymosin α with six histidine residues affixed to the carboxyl terminus (Trumbore et al. 1997), and pProβGN, which codes for a fusion protein containing 40 codons of Escherichia coli xanthine guanine phosphoribosyl transferase, 28 codons of E. coli tryptophanyl-tRNA synthetase, the entirety of prothymosin α, and E. coli β-galactosidase (Manrow et al. 1991) have already been described. A plasmid coding for prothymosin α with an N-terminal FLAG sequence, called pFLAG 60, was constructed by removing a Hind III-Bam HI fragment from pBC12BIProG, the prothymosin α gene behind the RSV LTR in an expression plasmid (Manrow et al. 1991), and replacing it with an engineered fragment. The unique Hind III site is located approximately 20 bases upstream of the transcriptional start site, whereas the Bam HI site is approximately 1.7 kb downstream in intron 1. A 1.5-kb Acc I-Bam HI fragment lacking the 20 bases of promoter sequence as well as most of exon 1 was obtained by cleaving the aforementioned Hind III-Bam HI fragment with Acc I. The region between Hind III and Acc I was rebuilt with oligomers that included a Hind III site for insertion, a Kozak sequence (Kozak 1986, 1987), and an initiator methionine codon followed by FLAG codons—asp-tyr-lys-asp-asp-asp-asp—adjacent to prothymosin α codons 2–6, i.e., ser-asp-ala-ala-and part of val. Because the prothymosin α codons were identical to those of the original gene, the cleaved Acc I site was preserved and the downstream prothymosin α region was fused in frame. After ligation to the Acc-Bam H1 fragment, the new Hind III-Bam HI fragment was inserted into the plasmid; the resultant construct consisted of a fused FLAG-prothymosin α gene with much of the 5′ untranslated region of the prothymosin α gene replaced with the new material. The clone was sequenced to ensure that it was correct.

Both COS-1 cells and NIH3T3 cells were transiently transfected. COS cells were plated in Lab-Tek double-well chamber slides (#178565; Nalge Nunc International, Naperville, IL) at 3 × 10⁴ cells per well and incubated overnight in complete medium. Transfection was accomplished with 250 ng of plasmid/well using LipofectAMINE (Life Technologies) and instructions provided by the manufacturer. Cells were allowed to express the ectopic histidine-tagged gene for 6 hr before experiment-specific manipulations were initiated, except in the studies described in Figure 7, where a new preparation of pBC12BIProG 6-his required ∼12 hr of expression to achieve similar product levels. The gene for prothymosin α fused to β-galactosidase was expressed for a minimum of 24 hr.

For NIH3T3 cells, transfections were performed in 60-mm dishes; cells were seeded at 4 × 10⁵ cells/dish and allowed to grow overnight, rinsed with DMEM, and transfected with 2 μg of plasmid using LipofectAMINE with PLUS reagent (Life Technologies) and a protocol provided by the manufacturer. After a 24–48-hr incubation to allow expression, cells were trypsinized (Manrow et al. 1991), replated in double-well chamber slides at 3 × 10⁴ cells/well, incubated overnight in complete medium, and fixed and stained as noted below.

Permeabilization of Cells

Permeabilization was carried out using minor modifications of the technique described by Adam et al. (1990). Briefly, ice-cold digitonin (#300410; Calbiochem, La Jolla, CA) at 40 μg/ml in transport buffer (20 mM HEPES at pH 7.3, 110 mM potassium acetate, 5 mM sodium acetate, 2 mM dithiothreitol, and 1 mM EGTA) was used to resuspend freshly trypsinized cells. After a 5-min incubation, the cells were layered on poly-l-lysine-coated microscope slides, carefully washed once with cold PBS, and fixed and stained as noted below.

Cells grown on microscope slides were permeabilized with α-toxin by bathing the monolayers in a solution of 300 hemolytic U/ml of α-toxin (#616385, Calbiochem) in transport buffer for 5 min on ice. Immediately thereafter, cells were incubated at 37C for 30 min in complete culture medium or complete medium containing 1 mM adenosine 5′-O-(3-thiotriphosphate) and 1 mM guanosine 5′-O-(3-thiotriphos-phate) (ATPγS and GTPγS) purchased from Sigma Chemical (St Louis, MO).

Immunostaining

Antibody treatments were performed on cells that were rinsed free of medium with PBS. Then the cells were fixed in PBS containing 3% paraformaldehyde in PBS for 30 min, permeabilized with 100% methanol, blocked overnight at 4C in blocking buffer [PBS containing 1% bovine serum albumin, 2% (v/v) non-immune horse serum, and 3% (v/v) non-immune goat serum], and stained with primary antibodies. Primary antibodies that recognized epitope tags on prothymosin α were mouse monoclonal antibody to β-galactosidase (#63363; ICN Biomedicals), mouse monoclonal antibody to pentahistidine (#34660; Qiagen, Valencia, CA), and anti-FLAG M2 antibody (Kodak; Rochester, NY). Other primary antibodies and their sources were as follows: mouse monoclonal antibody against the phosphorylated carboxyl terminal domain of RNA polymerase II, Covance (#MMS-134R; Richmond, CA); mouse monoclonal antibody against SC-35, BD PharMingen (#65201A; San Diego, CA); mouse monoclonal antibody against histone H1 (#382152; Calbiochem); and mouse monoclonal antibodies against karyopherin β, Ran, and transportin (# K48020, R32620, and T57720, respectively) from BD Transduction Laboratories (Lexington, KY). With the exception of the co-localization experiments with nuclear proteins (Figure 7), the secondary antibodies were either rhodamine-[tetramethylrhodamine isothiocyanate (TRITC)]-conjugated or fluorescein-[fluorescein isothiocyanate (FITC)]-conjugated AffiniPure goat anti-mouse IgG (Jackson ImmunoResearch Laboratories; West Grove, PA). For Figure 7, the dye-conjugated goat anti-mouse secondary antibodies were included in the Alexa Fluor 488 (green) or the Alexa Fluor 594 (red) Signal Amplification Kits (# 11054 and 11067, respectively; Molecular Probes, Eugene, OR). Samples were stained with antibodies diluted in blocking buffer for 1 hr at room temperature, washed three times for 10 min each with blocking buffer, and stained with diluted dye-conjugated antibodies. In some cases, tertiary antibodies, also supplied with the kits from Molecular Probes, were used to amplify weak signals. Coverslips were mounted using VECTASHIELD mounting medium with 4′, 6-diamidino-2-phenylindole (DAPI) (#H1200; Vector Laboratories, Burlingame, CA).

Microscopy and Image Analysis

Slides were viewed with a Leica DMRB fluorescence microscope (Leica; Rockleigh, NJ) and photographed using the SenSys camera system (Photometrics; Munich, Germany). The images were captured in gray scale and pseudocolored using the IP Lab Spectrum P software (Signal Analytics; Vienna, VA).

Confocal microscopy was performed with a Zeiss LSM 410 instrument using a krypton/argon laser for fluorescence excitation at 488 nm and 568 nm for FITC and TRITC, respectively. Images were captured using a pinhole setting of ∼1 AIRY unit for an optical section of 0.7 μm.

Final images were assembled in Adobe Photoshop 5.0.

Results

Location of Prothymosin α in Cells at Interphase and During M-phase

The location of prothymosin α in the cell was examined using in situ immunolocalization. COS-1 and NIH3T3 cells were transfected with genes for a series of epitopetagged prothymosin α proteins, shown diagramatically in Figure 1f, with the prothymosin α sequences colored green. Three variants bearing epitopes differing in size, charge, and position with respect to prothymosin α were generated as follows. Prothymosin α tagged with six histidine residues made use of a small basic epitope covalently attached to the carboxyl terminus (Figures 1a and 1b). Prothymosin α with FLAG employed an acidic epitope consisting of seven amino acids inserted immediately downstream of the amino-terminal methionine residue, which is cleaved from normal prothymosin α (Figure 1e). Finally, we used prothymosin α in a chimera consisting of 40 amino acids of E. coli xanthine-guanine phosphoribosyl transferase (Ecogpt), 28 amino acids from E. coli tryptophanyl-tRNA synthetase (trpS), and the entirety of prothymosin α fused to a very large bacterial protein without a nuclear localization signal of its own, i.e., β-galactosidase (Figures 1c and 1d). This molecule, abbreviated prothymosin α/β-gal, is bulky enough to be unable to diffuse passively in either direction through the nuclear pore. As shown in Figure 1, prothymosin α in NIH3T3 cells (left) or in COS-1 cells (right) was overwhelmingly nuclear regardless of whether the secondary antibody used for visualization was conjugated to rhodamine (red) or to fluorescein (green). It was also clear that the nuclear location of prothymosin α in the cell was independent of the nature of the epitope tag, its size, or its position within the expressed chimeric protein. In all cases, with the exception of the mitotic cell shown in Figure 1c, tagged prothymosin α appeared punctate but widely dispersed throughout the nucleus. When overexpressed, the tagged protein colored nuclei almost uniformly (data not shown), with the exception of the nucleolus, which remained essentially free of stained prothymosin α under all circumstances. The minor variations that were observed among these systems in Figure 1 can be ascribed to fluctuations in the intensity of the stain coupled with a range in transfection efficiencies and expression levels.

Expression of chimeras depends on the vector and the host. Transfected COS cells, which are capable of producing a huge excess of prothymosin α/6-his in 24–72 hr owing to the presence of large T-antigen in the cells and an SV40 origin of replication in the plasmid, generate detectable amounts only 2 hr after introduction of the DNA (data not shown). Our cells were examined 6 hr after transfection to limit the amount of tagged protein to about one quarter to one fifth that of its native counterpart. In contrast, FLAG/prothymosin α in NIH3T3 cells is synthesized at lower levels because the GC-rich region downstream of the TATA box was deleted. However, by 48 hr sufficient protein is accumulated in transfected cells to permeate the nucleus. Prothymosin α/β-gal expression suffers from an added deficiency. It makes use of a plasmid, pCH110, which not only lacks SV40 sequences but also contains a weaker promoter than that found in pBC12BI, the parent plasmid used for the FLAG and 6-his constructs. The robust expression of prothymosin α/6-his in COS cells, a transfection efficiency near 100%, and the ability to modulate the level of ectopic protein by varying the time of expression made this combination our preferred system in subsequent experiments. It is also worth noting that our laboratory has previously shown that prothymosin α/6-his is phosphorylated and dephos-phorylated normally in COS cells (Trumbore et al. 1997), an observation consistent with preservation of activity.

The one cell that differed qualitatively from all others in Figure 1 was a mitotic cell. Although some investigators have assumed that prothymosin α is released into the cytoplasm during mitosis, this is the first experimental illustration of the event. In Figure 1c, attached daughter cells in which cytokinesis and the reconstitution of the nuclei are virtually complete contained significant levels of cytoplasmic prothymosin α as well as a perinuclear accumulation of the protein, presumably poised for re-entry. To obtain further insight into the function of prothymosin α by exploring its location in cells under different conditions, we investigated COS cells throughout mitosis and compared the position of prothymosin α/6-his stained with rhodamine (Figure 2, Column 1) to that of DNA stained with DAPI (Figure 2, Column 2). The images are presented in mitotic order, with a cell in late prophase at the top, a cell in telophase at the bottom, and intermediate stages between the two. For ease in localization, the superimposed images of DNA and prothymosin α/6-his are shown in Figure 2, Column 3. The top two examples (Figure 2, Rows a and b) illustrate cells in interphase, which were lightly stained with DAPI, together with cells containing the condensed chromatin characteristic of prophase, which appeared much brighter in Figure 2, Column 2. In these interphase cells, the antibody for pentahistidine, which recognizes the chimeric prothymosin α, appeared to occupy the same area as the light DAPI stain (Figure 2, Column 2, Rows a and b). This perception was reinforced when the two images were overlaid (Figure 2, Column 3, Rows a and b). In contrast, in the cells in late prophase, prothymosin α appeared to be more diffuse than the brighter condensing DNA. When the images of prothymosin α and DNA were superimposed, it was dramatically clear in Figure 2, Row b (less so in row a) that a significant portion of prothymosin α was extranuclear well before the onset of prometaphase when the nuclear envelope breaks down. Later, in metaphase (Figure 2, Rows c and d), we were not surprised to find prothymosin α dispersed throughout the cell but virtually excluded from the metaphase plate. Throughout anaphase (Figure 2, Rows e–g), prothymosin α remained diffuse in the cell and continued to be excluded from the volume occupied by DNA. Finally, the location of prothymosin α was compared to that of chromatin at early telophase (Figure 2, Row h). Prothymosin α was vastly underrepresented in the region occupied by the newly forming nucleus. These data focus attention on the fact that prothymosin α can be found in the cytoplasm not only at mitosis, when the nuclear envelope ceases to exist, but also before the envelope breaks down and after it reforms. Therefore, prothymosin α is not exclusively nuclear.

Figure 1

Cellular distribution of epitope-tagged prothymosin α. NIH3T3 (a,c,e) or COS-1 (b,d) cells were transiently transfected with plasmids that direct the expression of chimeric prothymosin α proteins. The tagged proteins (f) are composed of the entirety of prothymosin α shown in green and a tag shown in white. The illustration diagrams the location of the tag and its size relative to prothymosin α, with the exception of the prothymosin α/β-gal construct, which contains 68 additional amino acids at the N-terminus and the entirety of β-gal (abbreviated in the diagram) at the C-terminus; numbers represent amino acid positions in the chimeric protein. The expressed proteins were identified with antibodies to pentahistidine, which recognizes the six C-terminal histidine residues (a,b), β-gal (c,d), or FLAG, which recognizes a seven amino-acid epitope (e). Secondary antibodies were conjugated to rhodamine (a,b) or to fluorescein (c-e).

Effect of Metabolic Inhibitors on the Distribution of Prothymosin α/6-his

The presence of prothymosin α in the cytoplasm of prometaphase cells raised questions about the location of the protein during interphase. Careful examination of cycling cells revealed labeled prothymosin α/6-his in the cytoplasm at levels too high to be attributed to background. Furthermore, the stain was not evenly dispersed throughout but appeared to be somewhat concentrated in a halo surrounding the nucleus (Figures 1b and 3a). Because protein synthesis is a cytoplasmic event, all newly synthesized proteins must traverse the cytoplasm en route to their final abode. We considered the possibility that cytoplasmic prothymosin α was no more than a reflection of this process. Accordingly, cells were transfected with prothymosin β/6-his DNA, incubated for 6 hr, and treated with cycloheximide (Figure 3b) or actinomycin D (Figure 3c) for 2 hr to shut down protein synthesis and RNA synthesis, respectively. As shown in Figure 3, these metabolic inhibitors, which themselves do not interfere with nuclear import (Gannon and Lane 1991; Pinol-Roma and Dreyfuss 1991), had no effect on the amount or the distribution of prothymosin α. Hence, prothymosin α molecules in the cytoplasm are not necessarily new. The possibility that a subset of prothymosin α is cytoplasmic or that the protein shuttles between the nuclear and cytoplasmic compartments had to be seriously considered.

Effect of Blocking Nuclear Import on the Distribution of Prothymosin α

Entry into the nucleus through the nuclear pore is a two-step process. In the first step, cytosolic proteins recognize and bind to the nuclear localization signal, forming a transport complex which docks with fibrils of the nuclear pore. In a second step, translocation through the pore complex occurs in an energy-dependent manner requiring the net hydrolysis of both ATP and GTP (reviewed in Ohno et al. 1998 and in Pemberton et al. 1998). To inhibit nuclear uptake of protein, transfected COS cells were cooled to 0C (Richardson et al. 1988; Breeuwer and Goldfarb 1990). Under these conditions, ectopic prothymosin α/6-his and other endogenous nuclear proteins can no longer transit to the nucleus. Three different proteins in two transfection experiments were examined. In the first transfection, prothymosin α/6-his DNA was introduced into COS cells and the product was visualized with rhodamine-conjugated secondary antibody (red), while endogenous Ran was examined with fluorescein-conjugated secondary antibody (green) in the same cells. In the second transfection, the much larger prothymosin α/β-gal, which was also visualized with rhodamine-conjugated secondary antibody, was the test protein (Figure 4). Ran was chosen as the endogeneous protein for comparison because it is small, it shuttles between nucleus and cytoplasm and, as a component of the nuclear transport apparatus, it must interact transiently with transport complexes bearing cargo proteins such as prothymosin α. At the initiation of the experiment when the cells remained at 37C, prothymosin α/6-his and Ran were predominantly massed in the nucleus, with only a trace of each in the cytoplasm. Prothymosin α/β-gal appeared to be virtually exclusively nuclear, as reported previously (Manrow et al. 1991) (Figure 1c and Figure 4, Column 1). After cooling of the cells for 2 hr, both prothymosin α/6-his and Ran moved into the cytoplasm in a time-dependent manner (Figure 4, Rows 1 and 2, Columns 2 and 3). In contrast, the prothymosin α/β-gal chimera was unaffected by the shift in temperature. No movement into the cytoplasm was detected and the tell-tale traces of cytoplasmic prothymosin α were absent (Figure 4, Row 3, Column 3). When the cooled cells were warmed to 37C, virtually all molecules of both prothymosin α/6-his and Ran reentered the nucleus within the 30-min test period (Figure 4, Column 4). In these experiments, Ran, which functions as a control protein, behaved as expected by accumulating in the cytoplasm under conditions in which entry into the nucleus was abrogated. Therefore, the data suggest that prothymosin α is also able to leave the nucleus and that its nuclear localization requires continuous re-entry. Furthermore, the data show that egress is not simply a function of prothymosin α sequences because both prothymosin α/6-his, which escapes from the nucleus, and prothymosin α/β-gal, which remains, have the prothymosin α moiety in common. We infer that the small size of native prothymosin α is essential for its mobility and that prothymosin α with an adduct the size of bacterial β-galactosidase passes unidirectionally into the nucleus.

Figure 2

Distribution of prothymosin α/6-his during mitosis. COS-1 cells were transiently transfected with a plasmid coding for histidine-tagged prothymosin α and stained with DAPI to highlight chromatin and with the antibody to pentahistidine followed by rhodamine-conjugated secondary antibody. Each row of photographs contains the same field illuminated for rhodamine fluorescence (Column 1), DAPI fluorescence (Column 2), and a superposition of the two images (Column 3).

Figure 3

Effect of inhibitors of RNA and protein synthesis on the distribution of prothymosin α. Transiently transfected COS-1 cells were fixed and stained for prothymosin α/6-his after growth under normal culture conditions (a), in the presence of cycloheximide (b), or in the presence of actinomycin D (c).

Figure 4

Distribution of tagged prothymosin α and Ran in chilled COS-1 cells. Stained cells were viewed under normal culture conditions at 37C (Column 1), after 1 hr at 0C (Column 2), after 2 hr at 0C (Column 3), and after 2 hr at 0C followed by 30 min at 37C (Column 4). The location of prothymosin α/6-his stained with rhodamine (Row 1), Ran stained with fluorescein (Row 2), and prothymosin α/β-gal also stained with rhodamine (Row 3) is displayed. For each condition, prothymosin α/6-his and Ran were examined in the same field of cells; the two examples of cells stained with antibody against β-gal came from a separate experiment.

We also investigated the fate of prothymosin α/6-his when nuclear import was prevented with a different technique. Cells were permeabilized with α-toxin (Bhakdi et al. 1993) and treated with ATPγS and GT-P7S to inhibit the energy-dependent component of protein uptake by the nucleus (Finlay et al. 1989; Melchior et al. 1993). In this format, the temperature of the cells remained at 37C but the integrity of the plasma membrane was violated to allow access of the nucleotide analogues. The experiment shows two representative cells. First, a control cell, which was treated with α-toxin but not with the analog inhibitors, was stained with DAPI (Figures 5a) and with antibodies to Ran (TRITC), and prothymosin α/6-his (FITC) (Figures 5b and 5c). Because the DAPI stain delineates the nucleus, it is clear that prothymosin α/6-his and Ran, although overwhelmingly nuclear, also populate the cytosol to a limited extent. Because this pattern has already been demonstrated in intact cells for both proteins (see Figure 4, Column 1), exposure to the toxin apparently did not affect the integrity of the nucleus or the preferred position of the two proteins. Under these conditions, it was possible to examine the effect of the analogue inhibitors in a second cell, shown in Figures 5d-5f. Here, both prothymosin α/6-his (Figure 5f) and Ran (Figure 5e) no longer reside almost exclusively in the nucleus. The FITC and TRITC labels, indicating the positions of the proteins, extend well beyond the volume occupied by the DAPI-stained DNA. Hence, a significant fraction of both proteins became cytosolic. The data corroborate the results of the cooling experiment: the nuclear localization of prothymosin α depends on repeated import into the nucleus. It is also obvious that the two proteins overlap in their distribution. This point becomes important, below, when the function of prothymosin α is addressed.

Figure 5

The effect of nonhydrolyzable nucleotide analogues on the distribution of prothymosin α/6-his (green) and Ran (red) in COS-1 cells permeabilized with Staphylococcus aureus α-toxin. Cells transiently expressing prothymosin α/6-his were permeabilized and incubated for 30 min in complete medium (a-c) or for 30 min in complete medium containing ATPγS and GTPγS (d-f). Cells were stained with DAPI (a,d), antibody against Ran (b,e), and antibody against pentahistidine (c,f). Each row shows the same cell with DAPI, TRITC, or FITC fluorescence on display.

The Distribution of Prothymosin α in Digitonin-treated Cells

The ability of prothymosin α molecules to diffuse rapidly throughout most of the nucleus and the cytoplasm could obscure a small confined pool of the protein located at its principal sites of activity. To identify such putative pools, we sought to deplete the nucleus of the mobile component of prothymosin α by treating the cells with digitonin. This technique, which perforates the plasma membrane, can remove almost all of the prothymosin α from a cell in a time- and dose-dependent manner. Digitonin treatment is also the preferred technique for generating systems for investigating nuclear transport, preparations known to be deficient in Ran (Moroianu and Blobel 1995). Accordingly, we compared Ran with prothymosin α in digitonin-treated cells. The results shown in Figure 6 present the same cell with FITC-stained Ran, TRITC-stained prothymosin α/6-his, and DNA stained with DAPI to define the boundaries of the nucleus. Although the particular cell chosen retained significant quantities of Ran in the nucleus, prothymosin α/6-his was completely absent; the photograph was overexposed so that the background red color would mark the nuclear outline. Despite the retention of Ran in the nucleus, which varied from nearly normal to barely detectable (data not shown), prothymosin α was quantitatively released. This experiment shows that prothymosin α is even more poorly anchored in the nucleus than Ran and suggests that neither prothymosin α in general nor a putative subset engages in tight binding to other subnuclear modules. Our data are consistent with the idea that the mobility of prothymosin α is important for its activity and that the domain in which it functions encompasses both the nucleus and the nearby surrounding cytosol.

Comparison of the Nuclear Position of Prothymosin α with That of Other Nuclear Proteins

Although our data were not consistent with the stable association of prothymosin α with any macromolecule, we nevertheless sought to discover whether it co-localized with well characterized nuclear molecules. To this end, we acquired a panel of antibodies that recognized proteins with different functions and different positions within the nucleus and we compared the location of histidine-tagged prothymosin α, stained with TRITC, with that of the selected second protein stained with FITC. We chose histone H1 for comparison because Gómez-Márquez and Rodríguez (1998) proposed that prothymosin α reorganizes chromatin into a more active configuration by binding to and depleting histone H1. This view is supported by two studies that demonstrated the binding of the two proteins in vitro (Papmarcaki and Tsolas 1994; Díaz-Jullien et al. 1996). In Figure 7a, it can be seen that both prothymosin α/6-his and histone H1 were broadly distributed in the nucleus and that histone H1 was extremely abundant. It is also clear that only a fraction of the two proteins co-localized, as indicated by a dearth of yellow denoting close juxtaposition of one with the other. Furthermore, when the image was manipulated to remove a wedge of histone H1 (FITC) in order to emphasize the TRITC stain of tagged prothymosin α, it became obvious that histone H1 extended to the periphery of the nucleus, whereas prothymosin α was concentrated more centrally. The image makes clear that a surfeit of green color had not simply obliterated the scarcer prothymosin α in red. These images are difficult to reconcile with a functional interaction. Presumably, even a transient interaction would require the distribution of the peripheral histone H1 to overlap with that of prothymosin α.

Figure 6

Effect of digitonin treatment on the distribution of prothymosin α/6-his and Ran in the same cell. Transfected COS-1 cells were permeabilized with digitonin and stained with DAPI (a), antibody against Ran (FITC) (b), and antibody against the histidine tag of prothymosin α (TRITC) (c).

Using an antibody to SC-35, we investigated the location of prothymosin α/6-his with respect to that of splicing factors in regions microscopically characterized as “speckled domains.” It is largely accepted that the speckled domains constitute storage depots for splicing factors; there is also evidence that these domains include staging areas for RNAs destined for export (reviewed in Spector 1993; Spector 1996). Because prothymosin α binds to the leucine motif/activation domains of HTLV-I Rex and HIV-1 Rev in vitro (Kubota et al. 1995), and because these viral proteins usher incompletely spliced messenger RNAs out of the nucleus, it seemed reasonable to seek prothymosin α in the vicinity of the speckled domains. As shown in Figure 7b, there was clearly a restricted distribution of SC-35 protein in the nucleus, as previously noted. More importantly, prothymosin α appeared to be abundant only in the regions separating the SC-35-stained domains. Not only were the speckled domains virtually free of prothymosin α but also there was no accumulation of prothymosin α at their perimeters. Such data do not support a direct role for prothymosin α in the recruitment or assembly of splicing factors.

The next pairwise comparison was carried out with antibodies against the phosphorylated C-terminal domain of RNA polymerase II (Bregman et al. 1995). We investigated an active RNA polymerase as a possible neighbor because the turnover of the high-energy glutamyl phosphates of prothymosin α is curtailed during mitosis and also in the presence of actinomycin D in NIH3T3 cells (Wang et al. 1997; Tao et al. 1999). At stages of the cell cycle other than mitosis, the phosphates remain labile. These data suggested a correlation between the activity of prothymosin α and ongoing transcription. At first glance, Figure 7c appears to corroborate this view. The tagged prothymosin α (TRITC) was much more plentiful than the active polymerase II molecules illuminated by the FITC-conjugated antibody; hence, there was widely distributed rhodamine fluorescence in the nucleus. In contrast, there was almost no green fluorescein fluorescence; the vast majority of polymerase II molecules were yellow, indicating co-localization with prothymosin α. Although one might infer a role for prothymosin α in the synthesis of mRNA from such co-localization data, the images to be presented below require a different, broader interpretation. A more global view of prothymosin α is also supported by quantitative assessments: there are ∼10⁵ growing transcripts of RNA in a HeLa cell (Jackson et al. 1998), whereas the estimates of native prothymosin α prevalence in a myeloma cell probably exceed 30 × 10⁶ (Sburlati et al. 1993).

Figure 7

The location of prothymosin α/6-his (TRITC-stained) in transiently transfected COS1 cells compared with the location of other nuclear proteins (FITC-stained). Cells were subjected to antibody against the histidine tail of prothymosin α (a-f) and to antibodies against histone H1 (a), SC-35 (b), the phosphorylated C-terminal domain of RNA polymerase II (c), Ran (d), transportin (e), or karyopherin β (f). Co-localization of prothymosin α with one of the other nuclear proteins is indicated (yellow). In a, a sector of FITC fluorescence was electronically removed to allow better viewing of the TRITC stain.

In Figure 7d-7f, the location of prothymosin α was compared pairwise with that of Ran, transportin, or karyopherin β, respectively. Each of these proteins is abundant, involved with transport through the nuclear pore, highly mobile (reviewed in Ohno et al. 1998 and in Pemberton et al. 1998) and, to a greater or lesser extent, each co-localized with prothymosin α. Ran, which is more abundant that the tagged prothymosin α but fairly equivalent in amount to the native prothymosin α in the cell, appeared mostly isolated from prothymosin α, but a significant portion was yellow. Prothymosin α was represented by rhodamine speckles and also by yellow, a clear indication that a part, but not the totality, of the two populations co-localized (Figure 7d). In contrast, co-localization of prothymosin α (red) with either transportin (Figure 7e) or karyopherin β (Figure 7f) was extensive, with a rhodamine tinge in Figure 7e denoting a slight excess of stained prothymosin α, whereas the fluorescein cast in Figure 7f was consistent with a tiny excess of stained karyopherin β. The data indicate that prothymosin α is found close to an array of molecules that depend on mobility to execute their functions.

Discussion

Our study explores the position of epitope-tagged prothymosin α in transfected cells as a means of directing attention towards processes in which the native protein may participate. Our data strongly suggest that prothymosin α, together with other highly diffusible molecules, resides predominantly in the nucleus in a domain characterized by internal fluidity, that there is measurable efflux into the cytosol, and that prothymosin α maintains its nuclear localization by continuous uptake through the nuclear pore. These conclusions are supported by a number of findings.

In lysed, disrupted, or leaky cells, prothymosin α is released into the cytoplasm. The ability to recover prothymosin α in post-nuclear supernatants of detergentlysed cells was demonstrated in 1990 by Sburlati et al., who first determined the abundance of the protein and later quantitatively recovered it in the soluble fractions of the cell. Similarly, digitonin treatment, which primarily affects the integrity of the plasma membrane, resulted in the depletion of tagged and native prothymosin α from all nuclei and a somewhat variable loss of Ran, a small GTP-binding protein required for transport of macromolecules through the nuclear pore. In contrast, cells treated with cytochalasin to solubilize actin filaments and centrifuged to cause enucleation retain prothymosin α within their nuclei. A fragment of plasma membrane seals each nucleus and prevents leakage of small molecules, including prothymosin α (Manrow et al. 1991). Therefore, from a biochemical standpoint, it is clear that prothymosin α is sufficiently mobile to diffuse rapidly from its initial location in the nucleus throughout the liquid phase unless plasma membranes intervene.

Prothymosin α is almost entirely nuclear in normal cells but exhibits a small, reproducible cytoplasmic presence. Using in situ immunolocalization techniques, together with specific antibodies that recognize both large and small epitope tags on prothymosin α, we have localized tagged prothymosin α primarily to the nucleus. However, further examination of our preparations revealed significant quantities of histidine-tagged prothymosin α in the cytosol in prometaphase cells as well as those in late telophase; a bias towards a perinuclear distribution was evident. Careful consideration of intact interphase cells using the same reagents confirmed the existence of extranuclear, small, tagged prothymosin α, exemplified by the histidine-tagged molecules. The cytoplasmic molecules persisted in the presence of actinomycin D and, more importantly, cycloheximide, a finding suggesting that newly synthesized prothymosin α molecules en route to a nuclear domain as well as mature molecules constitute the cytoplasmic pool. Hence, the data validate the presence of prothymosin α outside the nucleus.

In metabolically compromised cells, the cytoplasmic pool of prothymosin α becomes significant. Two techniques that abrogate nuclear uptake of proteins—cooling cells to 0C and blocking the energy-dependent step of transport with inhibitory levels of ATP7S and GTP7S intracellularly (Richardson et al. 1988; Breeuwer and Goldfarb 1990; Melchior et al. 1993)—caused histidine-tagged prothymosin α to accumulate in the cytoplasm; prothymosin α/β-gal, evidently, was too large to escape from the nucleus. These data strongly suggest that the ability of the nucleus to concentrate native prothymosin α depends on the protein's potent nuclear localization signal (Manrow et al. 1991) and on a functioning nuclear import system.

Finally, two further observations support our view of prothymosin α as an unfettered, mobile molecule. Attempts to immunoprecipitate prothymosin α tagged with the FLAG epitope in the presence of a high concentration of calcium did not succeed in identifying specific binding partners (Manrow, unpublished data). Moreover, attempts to crosslink reactive forms of prothymosin α and histidine-tagged prothymosin α in situ failed to distinguish neighboring molecules. In the most extensive of these experiments, histidine-tagged prothymosin α was covalently attached to sulfosuccinimidyl 2-[7-azido-4-methylcoumarin-3-acetamide]ethyl-1,3′-dithiopropinate (SAED) in the dark, the compound was forced into cells by electroporation, and the reactive moiety was light-activated only after large quantities of the construct had attained the nucleus. No crosslinked partners were exposed (R.-H. Wang and S. L. Berger, unpublished data). Hence, the biochemical evidence and the imaging studies document the extraordinary ability of prothymosin α to disperse throughout much of the nucleus, to exit into the perinuclear space under normal conditions, and to accumulate in the cytosol when nuclear re-entry is obstructed.

Few proteins are known to travel freely through the nucleus. Because all proteins arise in the cytoplasm, they evidently enter the nucleus with or without an escort and proceed to the sites at which they function. Mounting evidence suggests that signal sequences direct proteins to find and associate with subnuclear compartments (reviewed in Spector 1993 and in Leonhardt and Cardoso 1995), while binding domains that recognize DNA, RNA, or other nuclear proteins maintain the localization. Order within the membrane-less interior of the nucleus is imposed by structures such as the nuclear matrix or domains of chromatin, which act as nucleation foci for the assembly of huge macromolecular complexes. Although the elegant work of Lawrence and colleagues (Carter et al. 1993; Xing et al. 1993) showed that mRNAs destined for export traverse curvilinear tracks from the gene to the periphery of the nucleus, the mechanism by which macromolecules negotiate a fibrous interior network has not been widely investigated. Specifically, it is not known whether a diffuse-and-bind model is sufficient to explain correct localization or whether macromolecules supply energy for the journey or assist in the directional passage of proteins and nucleic acids through the nucleus. Our study has focused attention on this issue by emphasizing a truly soluble protein, prothymosin α, which shares a soluble domain with karyopherin β, transportin, and Ran.

Our experiments show that a subset of prothymosin α co-localized with active RNA polymerase II. This finding is important because we have postulated a role for prothymosin α in transcription based on the ability of actinomycin D to obstruct turnover of prothymosin α's glutamyl phosphates in NIH3T3 cells (Tao et al. 1999). We now refine our views. In this study, actinomycin D treatment of COS-1 cells had no effect on the distribution of tagged prothymosin α, yet the drug is known to cause structural changes in the nucleus, including chromatin remodeling, coalescence of speckled domains into larger clusters, and redistribution of transcriptionally active polymerase II (Bregman et al. 1995). Because histidine-tagged prothymosin α remained unchanged despite alterations in other components of the transcriptional machinery, we infer that the role of prothymosin α in transcription is ancillary. Furthermore, with a concentration of native prothymosin α hundreds of times higher than that of polymerase II in a normal nucleus, the data suggest that prothymosin α has responsibilities that extend well beyond the polymerization of RNA.

Here we propose a model of the nucleus that includes prothymosin α. However, before demonstrating how prothymosin α's characteristics can be incorporated into nuclear activities, it might be useful to eliminate putative functions that are not consistent with the data in this communication. We have found that prothymosin α does not co-localize with histone H1 located near the nuclear membrane and co-localizes poorly with chromatin. Hence, functions including nucleosome assembly and chromatin remodeling (Díaz-Jullien et al. 1996; Gómez-Márquez and Rodríguez 1998), which require an interaction with any histone, are made problematic by our imaging data. Furthermore, it is also unlikely that prothymosin α is involved in splicing; the speckled domains illuminated with antibodies against SC-35 (Spector et al. 1991), the so-called storage sites for splicing components, did not co-localize with tagged prothymosin α. Similarly, the extreme fluidity of intracellular prothymosin α is difficult to reconcile with the observed in vitro binding of prothymosin α via its amino groups to peptides representing the conserved leucine-motif/activation domains of the viral proteins HIV rev and HTLV rex. Our data strongly suggest that stable binding is not a characteristic of prothymosin α regardless of the putative binding partner.

Any model that purports to explain the function of prothymosin α must account for the apparent paradox of a highly negatively charged protein that becomes transiently even more negatively charged by the acquisition of high-energy glutamyl phosphates while remaining mobile in the nucleus. The model must also accommodate an abundant molecule with an unfolded conformation. Indeed, the very floppiness of prothymosin α implies broad specificity rather than a lock-and-key fit with a limited number of partners. Therefore, we suggest that the mobility of prothymosin α is its function and that this attribute acquires meaning in the context of an intact nucleus where, deep within interior crevices, electrostatic barriers to free movement of large and small molecules abound. We propose that prothymosin α maintains the flow of molecules by repelling negatively charged molecules and by binding to positively charged ones. Furthermore, because stable binding to any macromolecule apparently does not occur, we invoke the known rapid turnover of glutamyl phosphates (Wang et al. 1997) as the means by which prothymosin α resists becoming electrostatically trapped. Although such a mechanism is far from proven, it is tempting to speculate that high-energy glutamyl phosphates are either hydrolyzed to allow movement or perhaps are transiently transferred to positively charged groups to promote a more neutral environment. Restated, we believe that prothymosin α facilitates movement not only of itself but also of other molecules into and within the nucleus, particularly in highly charged environments. If one thinks of a finished nucleus with basic proteins bound to RNAs and basic histones associated with DNA, one might doubt the need for a facilitator molecule to overcome electrostatic interactions. With a plethora of competing charged moieties in the nucleus, our approach is to wonder how those very highly charged molecules get together correctly in the first place.

Footnotes

Acknowledgements

We thank Richard E. Manrow of the National Cancer Institute for the use of pFLAG60, which he fabricated while employed as a member of this laboratory. We would also like to thank Dr Carolyn L. Smith of the National Institute of Neurological Diseases and Stroke and Dr John A. Hanover of the National Insititute of Diabetes and Digestive and Kidney Diseases for their help on separate occasions with confocal microscopy.

References

Adam

Marr

Gerace

(1990) Nuclear protein import in permeabilized mammalian cells requires soluble cytoplasmic factors. J Cell Biol 111: 807–816

Auger

Stahli

Fabien

Monier

J-C

(1987) Intracellular localization of thymosin alpha 1 by immunoelectron microscopy using a monoclonal antibody. J Histochem Cytochem 35: 181–187

Bhakdi

Weller

Walev

Martin

Jonas

Palmer

(1993) A guide to the use of pore-forming toxins for controlled permeabilization of cell membranes. Med Microbiol Immunol (Berl) 182: 167–175

Breeuwer

Goldfarb

(1990) Facilitated nuclear transport of histone H1 and other small nucleophilic proteins. Cell 60: 999–1008

Bregman

van der Zee

Warren

(1995) Transcription-dependent redistributionof the large subunit of RNA polymerase II to discrete nuclear domains. J Cell Biol 129: 287–298

Bustelo

Otero

Gómez-Márquez

Freire

(1991) Expression of the rat prothymosin alpha gene during T-lymphocyte proliferation and liver regeneration. J Biol Chem 266: 1443–1447

Carter

Bowman

Carrington

Fogarty

McNeil

Fay

Lawrence

(1993) A three-dimensional view of precursor messenger RNA metabolism within the mammalian nucleus. Science 259: 1330–1335

Clinton

Graeve

El-Dorry

Rodriguez-Boulan

Horecker

(1991) Evidence for nuclear targeting of prothymosin and parathymosin synthesized in situ. Proc Natl Acad Sci USA 88: 6608–6612

Conteas

Mutchnick

Palmer

Weller

Luk

Naylor

Erdos

Goldstein

Panneerselvam

Horecker

(1990) Cellular levels of thymosin immunoreactive peptides are linked to proliferative events: evidence for a nuclear site of action. Proc Natl Acad Sci USA 87: 3269–3273

10.

Díaz-Jullien

Pérez-Estévez

Covelo

Freire

(1996) Prothymosin alpha binds histones in vitro and shows activity in nucleo-some assembly assay. Biochim Biophys Acta 1296: 219–227

11.

Enkemann

Pavur

Ryazanov

Berger

(1999) Does prothymosin alpha affect the phosphorylation of elongation factor 2? J Biol Chem 274: 18644–18650

12.

Eschenfeldt

Berger

(1986) The human prothymosin alpha gene is polymorphic and induced upon growth stimulation: evidence using a cloned cDNA. Proc Natl Acad Sci USA 83: 9403–9407

13.

Eschenfeldt

Manrow

Krug

Berger

(1989) Isolation and partial sequencing of the human prothymosin alpha gene family. J Biol Chem 264: 7546–7555

14.

Finlay

Newmeyer

Hartl

Horecka

Forbes

(1989) Nuclear transport in vitro. J Cell Sci Suppl 11: 225–242

15.

Fraga

García-Caballero

Domínguez

Pérez-Becerra

Beiras

Forteza

(1993) Immunohistochemical location of prothymosin α in regenerating human hepatocytes and hepatocellular carcinomas. Virchows Arch [A] 423: 449–452

16.

Frangou–Lazaridis

Clinton

Goodall

Horecker

(1988) Prothymosin alpha and parathymosin: amino acid sequences deduced from the cloned rat spleen cDNAs. Arch Biochem Biophys 263: 305–310

17.

Gannon

Lane

(1991) Protein synthesis required to anchor a mutant p53 protein which is temperature-sensitive for nuclear transport. Nature 349: 802–806

18.

Gast

Damaschun

Eckert

Schulze-Forster

Maurer

Müller-Frohne

Zirwer

Czarnecki

Damaschun

(1995) Prothymosin alpha: a biologically active protein with random coil conformation. Biochemistry 34: 13211–13218

19.

Goldstein

Low

TLK

McAdoo

McClure

Thurman

Rossio

Lai

C-Y

Chang

Wang

Harvey

Ramel

Meienhofer

(1977) Thymosin α1: isolation and sequence analysis of an immunologically active thymic polypeptide. Proc Natl Acad Sci USA 74: 725–729

20.

Gómez-Márquez

Rodríguez

(1998) prothymosin alpha is a chromatin-remodelling protein in mammalian cells. Biochem J 333: 1–3

21.

Gómez-Márquez

Segade

Dosil

Pichel

Bustelo

Freire

(1989) The expression of prothymosin alpha gene in T lymphocytes and leukemic lymphoid cells is tied to lymphocyte proliferation. J Biol Chem 264: 8451–8454

22.

Goodall

Domínguez

Horecker

(1986) Molecular cloning of cDNA for human prothymosin alpha. Proc Natl Acad Sci USA 83: 8926–8928

23.

Haritos

Blacher

Stein

Caldarella

Horecker

(1985) Primary structure of rat thymus prothymosin alpha. Proc Natl Acad Sci USA 82: 343–346

24.

Haritos

Goodall

Horecker

(1984) Prothymosin alpha: isolation and properties of the major immunoreactive form of thymosin α₁ in rat thymus. Proc Natl Acad Sci USA 81: 1008–1011

25.

Jackson

Iborra

Manders

Cook

(1998) Numbers and organization of RNA polymerases, nascent transcripts, and transcription units in HeLa nuclei. Mol Biol Cell 9: 1523–1536

26.

Karetsou

Sandaltzopoulos

Frangou-Lazaridis

Lai

C-Y

Tsolas

Becker

Papamarcaki

(1998) Prothymosin alpha modulates the interaction of histone H1 with chromatin. Nucleic Acids Res 26: 3111–3118

27.

Kozak

(1986) Point mutations define a sequence flanking the AUG initiator codon that modulates translation by eukaryotic ribosomes. Cell 44: 283–292

28.

Kozak

(1987) At least six nucleotides preceding the AUG initiator codon enhance translation in mammalian cells. J Mol Biol 196: 947–950

29.

Kubota

Adachi

Copeland

Oroszlan

(1995) Binding of human prothymosin alpha to the leucine-motif/activation domains of HTLV-I rex and HIV-1 rev. Eur J Biochem 233: 48–54

30.

Leonhardt

Cardoso

(1995) Targeting and association of proteins with functional domains in the nucleus: the insoluble solution. Int Rev Cytol 162B:303–335

31.

Makarova

Grebenshikov

Egorov

Vartapetian

Bogdanov

(1989) Prothymosin alpha is an evolutionary conserved protein covalently linked to a small RNA. FEBS Lett 257: 247–250

32.

Manrow

Sburlati

Hanover

Berger

(1991) Nuclear targeting of prothymosin alpha. J Biol Chem 266: 3916–3924

33.

Melchior

Paschal

Evans

Gerace

(1993) Inhibition of nuclear protein import by nonhydrolyzable analogues of GTP and identification of the small GTPase, Ran/TC4, as an essential transport factor. J Cell Biol 123: 1649–1659

34.

Michalewsky

Gabriel

Winter

Makofske

Danho

Shively

Biemann

Meienhofer

(1983) Confirmation of the primary structure of thymosin α₁ by microsequence analysis of limited acid and enzymatic hydrolysis fragments. J Pept Prot Res 21: 93–99

35.

Moroianu

Blobel

(1995) Protein export from the nucleus requires the GTPase, Ran, and GTP hydrolysis. Proc Natl Acad Sci USA 92: 4318–4322

36.

Oates

Erdos

(1989) Biochemical identification of thymosin alpha-1: its phylogenetic distribution and evolutionary implications. Comp Biochem Physiol 94B:759–763

37.

Ohno

Fornerod

Mattaj

(1998) Nucleocytoplasmic transport: the last 200 nanometers. Cell 92: 327–336

38.

Palvimo

Linnala-Kankkunen

(1990) Identification of a low-Mr acidic nuclear protein as prothymosin alpha. FEBS Lett 277: 257–260

39.

Panneerselvam

Haritos

Caldarella

Horecker

(1987) Prothymosin alpha in human blood. Proc Natl Acad Sci USA 48: 4465–4469

40.

Panneerselvam

Wellner

Horecker

(1988) The amino acid sequence of bovine thymus prothymosin alpha. Arch Biochem Biophys 250: 197–201

41.

Papmarcaki

Tsolas

(1994) Prothymosin alpha binds to histone H1 in vitro. FEBS Lett 345: 71–75

42.

Pemberton

Blobel

Rosenblum

(1998) Transport routes through the nuclear pore complex. Curr Opin Cell Biol 10: 392–399

43.

Pinol-Roma

Dreyfuss

(1991) Transcription-dependent and transcription-independent nuclear transport of hnRNP proteins. Science 253: 312–314

44.

Richardson

Mills

Dilworth

Laskey

Dingwall

(1988) Nuclear protein migration involves two steps: rapid binding at the nuclear envelope followed by slower translocation through nuclear pores. Cell 52: 655–664

45.

Sburlati

De La Rosa

Batey

Kurys

Manrow

Pannell

Martin

Sheeley

Berger

(1993) Phosphorylation of human and bovine prothymosin alpha in vivo. Biochemistry 32: 4587–4596

46.

Sburlati

Manrow

Berger

(1990) Human prothymosin alpha: purification of a highly acidic nuclear protein by means of a phenol extraction. Protein Express Purif 1: 184–190

47.

Schmidt

Werner

(1991) Nucleotide sequence of the murine prothymosin alpha cDNA and its deduced primary and secondary protein structure. Biochim Biophys Acta 1088: 442–444

48.

Spector

(1993) Macromolecular domains within the cell nucleus. Annu Rev Cell Biol 9: 265–315

49.

Spector

(1996) Nuclear organization and gene expression. Exp Cell Res 229: 189–197

50.

Spector

X-D

Maniatis

(1991) Associations between distinct premRNA splicing components and the cell nucleus. EMBO J 10: 3467–3481

51.

Szabo

Weksler

(1992) Is thymosin α₁ a thymic hormone? Clin Immunol Immunopathol 65: 195–200

52.

Tao

Wang

R-H

Enkemann

Trumbore

Berger

(1999) Metabolic regulation of protein-bound glutamyl phosphates: insights into the function of prothymosin alpha. J Cell Physiol 178: 154–163

53.

Trumbore

Manrow

Berger

(1998) Prothymosin alpha is not found in yeast. Prot Express Purif 13: 383–388

54.

Trumbore

Wang

R-H

Enkemann

Berger

(1997) Prothymosin alpha in vivo contains phosphorylated glutamic acid residues. J Biol Chem 272: 26394–26404

55.

Tsitsiloni

Yialouris

Sekeri-Pataryas

Haritos

(1989) Prothymosin alpha is not a nuclear polypeptide. Experientia 45: 332–334

56.

Tsitsiloni

Stiakakis

Koutselinis

Gogas

Markopoulos

Yialouris

Bekris

Panoussopoulos

Kiortsis

Voelter

Haritos

(1993) Expression of α-thymosins in human tissues in normal and abnormal growth. Proc Natl Acad Sci USA 90: 9504–9507

57.

Vartapetian

Makarova

Koonin

Agol

Bogdanov

(1988) Small cytoplasmic RNA from mouse cells covalently linked to a protein. FEBS Lett 232: 35–38

58.

Vega

Vidal

Hellman

Wernstedt

Domínguez

(1998) Prothymosin alpha stimulates Ca²⁺-dependent phosphorylation of elongation factor 2 in cellular extracts. J Biol Chem 273: 10147–10152

59.

Wang

R-H

Tao

Trumbore

Berger

(1997) Turnover of the acyl phosphates of human and murine prothymosin alpha in vivo. J Biol Chem 272: 26405–26412

60.

Watts

Cary

Crane-Robinson

(1989) Prothymosin alpha is a nuclear protein. FEBS Lett 245: 17–20

61.

Watts

Cary

Sautiere

Crane-Robinson

(1990) Thymosins: both nuclear and cytoplasmic proteins. Eur J Biochem 192: 643–651

62.

Xing

Johnson

Dobner

Lawrence

(1993) Higher level organization of individual gene transcription and RNA splicing. Science 259: 1326–1330