Formation and implications of a ternary complex of profilin, thymosin beta 4, and actin

Data from affinity chromatography, analytical ultracentrifugation, covalent cross-linking, and fluorescence anisotropy show that profilin, thymosin beta(4), and actin form a ternary complex. In contrast, steady-state assays measuring F-actin concentration are insensitive to the formation of such a complex. Experiments using a peptide that corresponds to the N terminus of thymosin beta(4) (residues 6-22) confirm the presence of an extensive binding surface between actin and thymosin beta(4), and explain why thymosin beta(4) and profilin can bind simultaneously to actin. Surprisingly, despite much lower affinity, the N-terminal thymosin beta(4) peptide has a very slow dissociation rate constant relative to the intact protein, consistent with a catalytic effect of the C terminus on conformational change occurring at the N terminus of thymosin beta(4). Intracellular concentrations of thymosin beta(4) and profilin may greatly exceed the equilibrium dissociation constant of the ternary complex, inconsistent with models showing sequential formation of complexes of profilin-actin or thymosin beta(4)-actin during dynamic remodeling of the actin cytoskeleton. The formation of a ternary complex results in a very large amplification mechanism by which profilin and thymosin beta(4) can sequester much more actin than is possible for either protein acting alone, providing an explanation for significant sequestration even if molecular crowding results in a very low critical concentration of actin in vivo.     

C-terminal variations in beta-thymosin family members specify functional differences in actin-binding properties

Mammalian cells express several isoforms of beta-thymosin, a major actin monomer sequestering factor, including thymosins beta4, beta10, and beta15. Differences in actin-binding properties of different beta-thymosin family members have not been investigated. We find that thymosin beta15 binds actin with a 2.4-fold higher affinity than does thymosin beta4. Mutational analysis was performed to determine the amino acid differences in thymosin beta15 that specify its increased actin-affinity. Previous work with thymosin beta4 identified an alpha-helical domain, as well as a conserved central motif, as crucial for actin binding. Mutational analysis confirms that these domains are also vital for actin binding in thymosin beta15, but that differences in these domains are not responsible for the variation in actin-binding properties between thymosins beta4 and beta15. Truncation of the unique C-terminal residues in thymosin beta15 inhibits actin binding, suggesting that this domain also has an important role in mediating actin-binding affinity. Replacement of the 10 C-terminal amino acids of thymosin beta15 with those of thymosin beta4 did, however, reduce the actin-binding affinity of the hybrid relative to thymosin beta15. Similarly, replacement of the thymosin beta4 C-terminal amino acids with those of thymosin beta15 led to increased actin binding. We conclude that functional differences between closely related beta-thymosin family members are, in part, specified by the C-terminal variability between these isoforms. Such differences may have consequences for situations where beta-thymosins are differentially expressed as in embryonic development and in cancer.     

Thymosin alpha 1-like peptides: localization and biochemical characterization in the rat brain and pituitary gland

Using a radioimmunoassay for thymosin alpha 1, endogenous thymosin-like peptides were characterized in the rat brain and pituitary gland. Thymosin alpha 1-like peptides were present in high concentrations in hypothalamus and pituitary extracts. These peptides were characterized using gel filtration techniques and the main peak of immunoreactive thymosin had a molecular weight similar to that of thymosin alpha 1 (3108 daltons). Using HPLC techniques, one main peak of immunoreactivity was present in brain extracts, whereas two peaks were present in pituitary extracts, one of which coeluted with thymosin alpha 1. The discrete regional distribution of thymosin alpha 1-like peptides was investigated and the highest densities of immunoreactive thymosin were present in the median eminence and arcuate nucleus of the hypothalamus, as well as the neurointermediate lobe of the pituitary. Due to the anatomical proximity of immunoreactive thymosin to loci containing known releasing factors and hormones, thymosin alpha 1-like peptides may function as neuroendocrine regulatory agents.     

Isolation and characterization of thymosin beta 9 Met from pork spleen

We have identified a new thymosin beta 4-like peptide in pork spleen. The new peptide (12 mg) and thymosin beta 4 (33 mg) were isolated from 230 g of spleen by solid phase extraction, preparative isoelectric focusing, and HPLC. The new peptide was termed thymosin beta 9 Met to indicate its close relationship to thymosin beta 9 from calf. The only difference from thymosin beta 9 is the substitution of leucine by methionine at position 6. This peptide replaces thymosin beta 10 which is the minor thymosin beta 4-like peptide in most mammals, e.g., in man, rat, mouse, cat, and rabbit. The structure was determined by amino acid analysis, tryptic digestion, and carboxypeptidase digestion. Pork spleen contains 192 micrograms of thymosin beta 4 and 117 micrograms of thymosin beta 9 Met per gram of tissue.     

One-step procedure for the determination of thymosin beta 4 in small tissue samples and its separation from other thymosin beta 4-like peptides by high-pressure liquid chromatography

Thymosin beta 4 has been determined by a simple and fast one-step procedure in different tissues of rats. The tissues (1 to 40 mg) were disintegrated and deproteinized by homogenization in perchloric acid. After neutralization by potassium hydroxide the supernatant solution was used for determining thymosin beta 4 by reverse-phase HPLC without further manipulations. Not only does this procedure avoid artificial proteolysis as effectively as extraction of tissues by guanidinium chloride or boiling buffer, but it offers two further advantages. First, no additional steps--as for example desalting--are necessary prior to HPLC and thus the risk of losing thymosin beta 4 is eliminated. Using this procedure thymosin beta 4 is recovered quantitatively. The method is linear over the range 0.04 to 1.13 nmol and thymosin beta 4 is well separated from other thymosin beta 4-like peptides known to be present in mammals; i.e., thymosin beta Ala4, thymosin beta 9, thymosin beta 10, and thymosin beta Arg10. Second, the acid-insoluble pellet of the same extract can be used to determine the DNA content of the sample. Thus it is possible to relate thymosin beta 4 to DNA, which then allows comparing cells of different tissues and cell lines to one another. This procedure is also applicable to small peptides soluble in perchloric acid.     

Production in Escherichia coli of human thymosin beta 4 as chimeric protein with human tumor necrosis factor

We have produced thymosin beta 4 protein in Escherichia coli as a chimeric protein with tumor necrosis factor (TNF). The protein was abundantly expressed, was immunoreactive against both anti-thymosin beta 4 and anti-TNF antibodies, and retained cytotoxicity in a TNF assay using mouse L929 fibroblasts. This latter characteristic enabled the easy and simple purification of thymosin beta 4 merely by following the TNF activity. The chimeric protein was designed to have an Asp-Pro bridge between thymosin beta 4 and TNF which could be specifically cleaved under suitable acidic conditions to release the thymosin beta 4 from the chimeric protein. These results indicate that the expression system in E.coli of a chimeric protein composed of thymosin beta 4 and TNF is appropriate for obtaining an abundant amount of the beta 4 peptide, especially since its purification from tissues is usually difficult because of limited yield and obscurity of its biological activity.     

Retinoic Acid Regulates Thymosin β Levels in Rat Neuroblastoma Cells

A small acidic polypeptide, termed thymosin beta 10, has been identified and is present in the nervous system of the rat by the ninth day of gestation. Thymosin beta 10 levels rise during the remaining days of life in utero, and then decline to nearly undetectable values between the second and fourth week post partum. The present study investigates the possible developmental signals and mechanisms that might regulate the expression of thymosin beta 10 during neuroembryogenesis. Many cell lines derived from tumors of the central nervous system express thymosin beta 10, as well as its homologue gene product, thymosin beta 4. Because some of these cell lines respond to exogenously applied agents by increasing their apparent state of differentiation, we have determined whether thymosin beta 10 levels are coordinately modulated. In several neuroblastomas, including the B103 and B104 lines, retinoic acid elicits a time- and dose-dependent increase in the content of thymosin beta 10, but not that of thymosin beta 4. The increase in thymosin beta 10 polypeptide is associated with a marked increase in the specific mRNA encoding this molecule. The mRNA for thymosin beta 4 is unaffected by retinoic acid. This is in contrast with the situation in vivo, where the expression of both genes decreases after birth. Other agents that influence the morphology of B104 cells, such as phorbol esters and dibutyryl cyclic AMP, have no influence on beta-thymosin levels. A range of steroids, which like retinoids act upon nuclear receptors, was also inactive. The stimulatory action of retinoic acid is detectable within 4 h, and thymosin beta 10 peptide levels continue to rise for at least 4 days. The influence of the isoprenoid is fully reversible and exhibits structural specificity. We believe that this culture system is mimicking the early rising phase of thymosin beta 10 levels in brain and that endogenous retinoids may be candidate physiological regulators of this gene.     

Synthesis of an artificial gene coding for thymosin alpha1 and its expression in Escherichia coli as a hybrid with human tumor necrosis factor]

Chemical-enzymatic synthesis and cloning in Escherichia coli of an artificial gene encoding the immunoactive peptide thymosin alpha 1 have been carried out. Recombinant plasmids were constructed which contain fusion genes coding for hybrids of human tumour necrosis factor (TNF) and thymosin alpha 1 as N- or C-terminal part of the hybrid protein. In the C-terminal hybrid protein, TNF and thymosin alpha 1 are linked through a methionine residue, thus allowing for thymosin alpha 1 to be cleaved off the rest of the hybrid protein with cyanogen bromide. In case of the N-terminal hybrid protein, the linker between the thymosin alpha 1 and TNF sequences is the acid-labile dipeptide Asp-Pro. Expression of the hybrid genes in E. coli and properties of the recombinant proteins were studied. The N-terminal hybrid protein was secreted into periplasmic space, in contrast with the C-terminal hybrid protein, which formed insoluble aggregates inside bacterial cells. Procedures for the isolation of both hybrid proteins were developed. The N-terminal hybrid protein displayed full biological activity in the cytotoxic assay on the mouse fibroblast L-929 whereas the C-terminal hybrid protein proved to be much less active. Treatment of the hybrid protein TNF-thymosin alpha 1 with cyanogen bromide lead to a mixture of two polypeptides, from which thymosin alpha 1 was purified to homogeneity by simple chromatographic procedures.     

Developmental Regulation of β-Thymosins in the Rat Central Nervous System

HPLC analysis of guanidinium hydrochloride extracts of neonatal and adult rat brain revealed a polypeptide that is present in high concentration in the immature nervous system, but whose levels decline dramatically in the adult. This polypeptide has been isolated and its complete amino acid sequence determined by gas-phase Edman degradation following specific chemical and enzymatic cleavages. The molecule is identified as thymosin beta 10, a member of a multigene family that encodes a structurally conserved series of small acidic polypeptides of uncertain function. Thymosin beta 10 is present in the developing nervous system as early as embryonic day 9. Levels subsequently increase to peak values between embryonic day 15 and postpartum day 3, before falling to adult values (about a 20-fold reduction) by postpartum day 14. The elevated levels of thymosin beta 10 in fetal and neonatal brain correlate with high levels of thymosin beta 10 mRNA, whereas the low values of the polypeptide in the adult and juvenile are mirrored by an approximate 15-fold reduction in specific mRNA. In comparison, the levels of thymosin beta 4 polypeptide, a homologue of thymosin beta 10, only decline by about 20% during the same developmental period. However, the mRNA encoding thymosin beta 4 is elevated in fetal brain, and its levels decrease approximately four-fold to a stable value around the time of birth. The reason for this discrepancy between thymosin beta 4 protein and mRNA levels is unknown. Thymosin beta 10 can also be detected by HPLC in fetal liver, where levels are approximately 5% of those in brain. In liver, thymosin beta 10 also declines following birth. It is concluded that beta-thymosin expression (as measured by steady-state mRNA and polypeptide levels) is both up- and down-regulated during different phases of maturation of the mammalian nervous system.     

Molecular cloning of the cDNA for rat spleen thymosin beta 10 and the deduced amino acid sequence

A rat spleen cDNA library was prepared and employed for the molecular cloning of the cDNA for thymosin beta 10, a peptide that previously had been found to accompany the closely related peptide, thymosin beta 4, in several species of mammals (S. Erickson-Viitanen, S. Ruggieri, P. Natalini, and B. L. Horecker (1983) Arch. Biochem. Biophys. 225, 407-413). First-round screening with a synthetic oligodeoxynucleotide probe yielded 55 positive clones, and sequence analysis of 11 of these clones revealed that they all coded for a peptide containing the thymosin beta 10 sequence, except for an additional arginyl residue at position 39. This peptide, designated thymosin beta 10arg, had been identified previously in rabbit tissues and reported as a variant of thymosin beta 10 (S. Ruggieri, S. Erickson-Viitanen, and B.L. Horecker (1983) Arch. Biochem. Biophys. 226, 388-392). Analysis of the 55 positive clones using a specific oligodeoxynucleotide probe constructed to correspond to the mRNA sequence, including the codon for Arg39, confirmed that they all coded for the amino acid sequence including Arg39. Based on these results, the existence of a molecular species lacking Arg39 is considered unlikely, and we conclude that thymosin beta 10 contains 43, rather than 42, amino acid residues, with identity to thymosin beta 4 in 32 of the 43 residues. We propose that the name thymosin beta 10 be used to refer to the peptide containing Arg39 and that the designation thymosin beta 10arg be dropped. In the cDNA sequence the codons for Ala1 and Ser43 of thymosin beta 10 are flanked by initiator and terminator codons, respectively; thus, both the thymosin beta 4 and thymosin beta 10, which coexist in mammalian cells and tissues, are synthesized without the formation of larger polypeptide precursors.     

Biosynthesis rates and content of thymosin beta 4 in cell lines

The content and relative biosynthetic rates of thymosin beta 4 have been determined in 28 different cell lines. The highest content of thymosin beta 4 as well as the highest rate of biosynthesis was observed in Epstein-Barr virus-transformed human B-cell lines. The levels observed in these cells are 1 pg thymosin beta 4 per cell, which is three times higher than that in rat peritoneal macrophages. Thus, these B-cell lines have the highest content of thymosin beta 4 of any cell type yet described. Since all of the Epstein-Barr virus-transformed B-cells described here grow in suspension, it is unlikely that the presence of thymosin beta 4 is related to anchorage in these cells. Thymosin beta 4 is not secreted by viable Epstein-Barr virus-transformed B cells in culture, suggesting some intracellular function of the peptide. These results indicate that these B-cell lines may be suitable for the study of thymosin beta 4 function.     

Cloning and characterization of a testis-specific thymosin beta 10 cDNA. Expression in post-meiotic male germ cells.

Thymosin beta 10 is one of a small family of proteins closely related in sequence to thymosin beta 4, recently identified as an actin-sequestering protein. A single molecular weight species of thymosin beta 10 mRNA is present in a number of rat tissues. In adult rat testis, an additional thymosin beta 10 mRNA of higher molecular weight was identified. Nucleotide sequencing of cDNA clones complementary to the testis-specific thymosin mRNA indicated that this mRNA differed from the ubiquitous thymosin beta 10 mRNA only in its 5'-untranslated region, beginning 14 nucleotides upstream of the translation initiation codon. These results, together with primer extension experiments, suggest that the two thymosin beta 10 mRNAs are transcribed from the same gene through a combination of differential promoter utilization and alternative splicing. The novel thymosin beta 10 mRNA could be detected only in RNA isolated from sexually mature rat testis. Both mRNAs were present in pachytene spermatocytes; only the testis-specific mRNA was detected in postmeiotic haploid spermatids. Immunoblot analysis using specific antibodies showed that the thymosin beta 10 protein synthesized in adult testis was identical in size to that synthesized in brain. Immunohistochemical analysis showed that the protein was present in differentiating spermatids, suggesting that the testis-specific thymosin beta 10 mRNA is translated in haploid male germ cells.     

Thymosin beta 4Xen: a new thymosin beta 4-like peptide in oocytes of Xenopus laevis

Two new thymosin beta 4-like peptides have been detected in ovaries of Xenopus laevis and Rana esculenta. Previously, it was reported that thymosin beta 4 can be found in various species, from mammals to amphibians, e.g., in X. laevis [S. Erickson-Viitanen, S. Ruggieri, P. Natalini, and B.L. Horecker (1983) Arch. Biochem. Biophys. 221, 570-576]. However, oocytes and spleen from R. esculenta contain no thymosin beta 4 but a similar peptide without methionine. The peptide from R. esculenta elutes from a reversed-phase column about 5 min later than thymosin beta 4. The peptide from X. laevis, referred to as thymosin beta 4Xen, can hardly be distinguished from thymosin beta 4 by its retention time on HPLC, by amino acid analysis, its isoelectric point, or tryptic fingerprinting. Amino acid analyses of the tryptic fragments, however, have revealed that thymosin beta 4 and beta 4Xen are different. The amino acid sequence of thymosin beta 4Xen is reported. Thymosin beta 4 and beta 4Xen differ in the amino acid residues at positions 15, 40, and 41. At position 15 serine is replaced by alanine and at 41-42 the sequence is Thr-Ser instead of Ala-Gly. Depending on their size, defolliculated oocytes contain between 2.7 and 52.6 ng thymosin beta 4Xen which is comparable to the amount of histones in oocytes.     

Thymosin beta10 inhibits cell migration and capillary-like tube formation of human coronary artery endothelial cells

Thymosin beta10 is a cytoplasm G-actin sequestering protein whose functions are largely unknown. To determine the direct effects of exogenous thymosin beta10 on angiogenic potentials as endothelial cell migration and capillary-like tube formation, human coronary artery endothelial cells (HCAECs) were incubated with increasing doses of thymosin beta10 (25-100 ng/ml). By using a modified Boyden chamber assay, thymosin beta10 inhibited cell migration in a dose- and time-dependent manner with the maximal effect being a 36% reduction at 100 ng/ml as compared to controls (P < 0.01). In addition, thymosin beta10 (100 ng/ml) significantly inhibited the capillary-like tube-formation of HCAECs on Matrigel, showing a 21% reduction of the total tube length as compared to negative controls (P < 0.01). Furthermore, by using real time PCR analysis, thymosin beta10 significantly decreased mRNA levels of vascular endothelial growth factor (VEGF), VEGF receptor-1 (VEGFR-1) and integrin alphaV after 24 h treatment in HCAECs. By contrast, thymosin beta4 significantly increased HCAEC migration. These results indicate that thymosin beta10, but not thymosin beta4, have direct inhibitive effects on endothelial migration and tube formation that might be mediated via downregulation of VEGF, VEGFR-1 and integrin alphaV in HCAECs. This study suggests a potential therapeutic application of thymosin beta10 to the diseases with excessive angiogenesis such as cancer.     

Immunohistochemical localization of thymosin beta 9 in bovine tissues

Using an indirect fluorescent antibody technique with frozen sections, the localization of thymosin beta 9 was investigated for the first time in bovine thymus, spleen, lung, muscle and liver. The antibodies used have been raised against the N-terminal fragment 1-14 of thymosin beta 9 in order to minimize the cross-reactivity with thymosin beta 4 which was found to be also present in bovine tissues. The specific antibodies against thymosin beta 9 raised in our laboratory allowed us to localize this peptide in presence of the highly homologous and always accompanying thymosin beta 4 in different tissues. Although thymosin beta 9 was first isolated from calf thymus, it could be also detected in other bovine organs. The highest density of positive immunoreaction was found to be in spleen sections. In the muscle tissue a pronounced fluorescence intensity was present in the region of the sarcolemma.     

Zyxin is upregulated in the nucleus by thymosin beta4 in SiHa cells

Thymosin beta4 is a 43-amino acid actin-binding protein that promotes cell migration and is important in angiogenesis, wound healing, and tumor metastasis. We searched for genes upregulated by thymosin beta4 and identified zyxin as increased in SiHa cells in the presence of exogenously added thymosin beta4 and when thymosin beta4 is overexpressed using adenoviral vectors. Both zyxin and thymosin beta4 show increased localization in the nucleus. We conclude that thymosin beta4 may exert some of its migration promoting activity via increased zyxin expression.     

Localization of thymosin β10 in breast cancer cells: relationship to actin cytoskeletal remodeling and cell motility

Beta-thymosins are polypeptides involved in the regulation of actin polymerization and thymosin β10 and β4 have been implicated in sequestration of monomeric (G-) actin. Additionally, experimental overexpression of thymosin β10 has been found to result in increases in F-actin bundles as well as in cell motility and spreading. We have studied the distribution of endogenously expressed thymosin β10 in cultured human breast cancer cell lines. Both unperturbed monolayer cultures and wound-healing models were examined using double-staining for thymosin β10 and polymerized (F-) actin. Our findings show that thymosin β10 is expressed in all three-cancer cell lines (SK-BR-3, MCF-7 and MDA-MB-231) studied. No or little staining was detected in confluent cells, whereas strong staining occurred in semiconfluent cells and in cells populating monolayer wounds. Importantly, the distribution of staining for thymosin β10 was inverse of staining for F-actin. These data support a physiological role for thymosin β10 in sequestration of G-actin as well as in cancer cell motility.     

The control of actin nucleotide exchange by thymosin beta 4 and profilin. A potential regulatory mechanism for actin polymerization in cells.

We present evidence for a new mechanism by which two major actin monomer binding proteins, thymosin beta 4 and profilin, may control the rate and the extent of actin polymerization in cells. Both proteins bind actin monomers transiently with a stoichiometry of 1:1. When bound to actin, thymosin beta 4 strongly inhibits the exchange of the nucleotide bound to actin by blocking its dissociation, while profilin catalytically promotes nucleotide exchange. Because both proteins exchange rapidly between actin molecules, low concentrations of profilin can overcome the inhibitory effects of high concentrations of thymosin beta 4 on the nucleotide exchange. These reactions may allow variations in profilin concentration (which may be regulated by membrane polyphosphoinositide metabolism) to control the ratio of ATP-actin to ADP-actin. Because ATP-actin subunits polymerize more readily than ADP-actin subunits, this ratio may play a key regulatory role in the assembly of cellular actin structures, particularly under circumstances of rapid filament turnover.     

Thymosin beta arg10, a major variant of thymosin beta 10 in rabbit tissues

Two homologous peptides, designated thymosin beta 4 and thymosin beta 10, respectively, have been shown to be widely distributed in mammalian cells and tissues (S. Erickson-Viitanen, S. Ruggieri, P. Natalini, and B.L. Horecker (1983) Arch. Biochem. Biophys. 221, 570-576; S. Erickson-Viitanen, S. Ruggieri, P. Natalini, and B.L. Horecker, (1983) Arch. Biochem. Biophys. 225, 407-413). In the rabbit, thymosin beta 4 is replaced by a variant, thymosin beta ala4, that contains alanine in place of serine at the blocked NH2-terminus. It is reported that in rabbit tissues thymosin beta 10 is also replaced by a variant, designated thymosin beta arg10, that contains an additional amino acid, arginine, inserted following lysine-38. The rabbit tissues analyzed also differ from those of other mammals in the relative quantities of thymosin beta ala4 and beta arg10, which are nearly equal, compared to tissues from other mammals where the quantities of thymosin beta 10 are only one-third to one-tenth those of thymosin beta 4.